CN103325121A - Method and system for estimating network topological relations of cameras in monitoring scenes - Google Patents

Abstract

The present invention is applicable to the field of security technology, and provides a method and system for estimating the topological relationship of a camera network in a monitoring scene. The method includes: decomposing the monitoring scene in the video stream captured in the monitoring network into grids; obtaining the The color histogram information of the optical flow of each grid in the monitoring scene; according to the color histogram information of the optical flow of each grid in the monitoring scene, the grids in the monitoring scene are clustered to obtain the monitoring scene The semantic region segmentation results of each monitoring scene; determine the network topology relationship between each camera according to the semantic region segmentation results of each monitoring scene. The present invention solves the problem existing in the prior art that the calculation of the topological relationship between cameras is based on the positioning and tracking of specific target activities, and the algorithm performance will drop sharply when there is occlusion in the monitoring environment or the resolution of the monitoring image is low .

Description

Method and system for estimating topology relationship of camera network in surveillance scene

technical field

The invention belongs to the field of security technology, and in particular relates to a method and system for estimating the topological relationship of a camera network in a monitoring scene.

Background technique

The topology estimation of the camera network is a key issue in the deployment of the camera network. Accurate topology estimation can not only grasp the movement patterns of people and crowds in the monitoring area, but also can further optimize the deployment through feedback.

Existing technologies provide some methods for topology estimation of camera networks, including:

1. Based on the results of people detection and tracking based on image background elimination, the out-of-control correlation of crowd activities among multiple cameras can be obtained to provide a basis for analyzing and establishing the target activity pattern of the entire scene.

2. Use the personnel pace information captured by multiple cameras to obtain the general mode of personnel activities, and readjust the camera deployment according to this mode, so as to reach the monitoring target with a better viewing angle and fewer cameras.

3. A hybrid probability density estimator based on the Parzen window and Gaussian kernel to estimate the probability density function composed of the time interval, the position of entering and exiting the observation field, and the movement speed when entering and exiting the field of view. The entire estimation process is passed through the learning training set Data method implementation.

Fourth, in terms of time domain constraints, a fuzzy time interval is used to represent the possibility of the observed target appearing in the next camera, and this possibility is estimated by the motion equation.

5. Using a large amount of target observation data, through the method of unsupervised learning, the spatio-temporal topological relationship between cameras is automatically established for a multi-camera surveillance network. On this basis, they also gave a method to verify the performance of the algorithm and realized the tracking of the target in the network.

Sixth, using a more general information theory idea about statistical trust, combining uncertainty correspondence and Bayesian method, reducing assumptions and showing better performance.

7. Assuming that all cameras have potential connections, and then remove the impossible connections through observations, the experiments prove that their method is effective in learning the topological relationship of large-scale camera networks, especially in the case of fewer learning samples. Better efficiency and effectiveness.

8. A lot of work utilizes the topological relationship of multiple cameras for global activity analysis and pedestrian re-identification.

However, the above topology inference algorithms are basically based on the positioning and tracking of specific target activities, and have high requirements for the quality of surveillance video. When there is occlusion in the surveillance environment or the resolution of the surveillance image is low, the performance of the algorithm will drop sharply.

Contents of the invention

The purpose of the embodiment of the present invention is to provide a method and system for estimating the topological relationship of the camera network in the monitoring scene, so as to solve the problems existing in the prior art. The existing topology inference algorithms are basically based on the positioning and tracking of specific target activities. The video quality requirements are high. When there is occlusion in the monitoring environment or the resolution of the monitoring image is low, the performance of the algorithm will drop sharply.

The embodiment of the present invention is achieved in this way, a method for estimating the topology relationship of a camera network in a surveillance scene, the method comprising the following steps:

Decompose the monitoring scene in the video stream captured by each camera in the monitoring network into a grid;

For each monitoring scene, obtain the color histogram information of the optical flow of each grid in the monitoring scene;

For each monitoring scene, cluster the grids in the monitoring scene according to the color histogram information of the optical flow of each grid in the monitoring scene, and obtain the semantic region segmentation result of the monitoring scene;

According to the semantic region segmentation results of each monitoring scene, the network topology relationship among the cameras in the monitoring network is determined.

The purpose of another embodiment of the present invention is to provide a camera network topology relationship estimation system in a surveillance scene, the system comprising:

A decomposition unit, configured to decompose the monitoring scene in the video stream captured by each camera in the monitoring network into grids;

An acquisition unit, for each monitoring scene, to obtain the color histogram information of the optical flow of each grid in the monitoring scene;

A clustering unit, for each monitoring scene, clustering the grids in the monitoring scene according to the color histogram information of the optical flow of each grid in the monitoring scene, to obtain the semantic area of the monitoring scene segmentation result;

The determining unit is configured to determine the network topology relationship among the cameras in the monitoring network according to the semantic region segmentation result of each monitoring scene.

The embodiment of the present invention calculates the color histogram feature of the optical flow of the grid through the optical flow algorithm, and further calculates the topological relationship between the cameras, and does not need to obtain the positioning or tracking of the moving target very clearly, which solves the problems in the prior art. Yes, the calculation of the topological relationship between cameras is based on the positioning and tracking of specific target activities, and the quality of the surveillance video is high. When there is occlusion in the surveillance environment or the resolution of the surveillance image is low, the performance of the algorithm will drop sharply. .

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the accompanying drawings that need to be used in the descriptions of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only of the present invention. For some embodiments, those of ordinary skill in the art can also obtain other drawings based on these drawings without any creative effort.

Fig. 1 is the implementation flowchart of the method for estimating the topology relationship of the camera network in the monitoring scene provided by an embodiment of the present invention;

FIG. 2 is a diagram of camera topology estimation results provided by another embodiment of the present invention;

Fig. 3 is a floor and camera deployment diagram provided by another embodiment of the present invention;

Fig. 4 is a block diagram of a system for estimating the topology relationship of a camera network in a surveillance scene according to another embodiment of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

An embodiment of the present invention provides a method for estimating the topology relationship of a camera network in a monitoring scene. The method is shown in FIG. 1, and the specific steps include:

In step S101 , the monitoring scene in the video stream captured by each camera in the monitoring network is decomposed into grids.

In this embodiment, at least two cameras are included in the monitoring network, and each camera captures a video stream. The video stream includes multiple frames, and each frame is an image. In multiple frames, a moving target passes by The image is the surveillance scene.

It should be noted that the grid size is generally 10*10, which can also be preset, but the decomposed grid size of the monitoring scene in the video stream captured by all cameras is the same.

In step S102, for each monitoring scene, the color histogram information of the optical flow of each grid in the monitoring scene is acquired.

Specifically, the method for obtaining the color histogram information of the optical flow of each grid in the monitoring scene is specifically as follows:

Define the video stream captured by the camera as I _n (X), where X is the coordinate of the grid in the monitoring scene of the video stream, X=(x; y) ^T , and the x is the horizontal axis of the grid Coordinates, the y is the ordinate of the grid, the T represents the transposition of the matrix, and the n is the number of the video frame included in the video stream;

definition $W(x;p)=\left(\begin{array}{c}(1+p1)\×x+p3\×\mathrm{the\; y}+p5\\ p2\×x+(1+p4)\×\mathrm{the\; y}+p6\end{array}\right);---\left(1\right)$

Wherein, W represents the variability template, p=(p1, p2, p3, p4, p5, p6) ^T , the p1, p2, p3, p4 are 0, and the p5, p6 are the optical flow information of the grid;

definition $p=\arg \underset{p}{\min }\underset{x}{\mathrm{\Σ}}{[I\left(W(x;p+\mathrm{\Δp})\right)-T\left(x\right)]}^2;---\left(2\right)$

Among them, Δp represents the difference of p between two iterations, and T(x) represents the decomposed grid of the first frame of the video stream;

It should be noted that the decomposed grid of the first frame of the video stream refers to the decomposed grid of the first frame image in the video stream.

Iterate according to (3), (4), and (5) until Δp is less than the preset threshold ε;

$<span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&Delta;x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\begin{array}{c}\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; Ix</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}\\ \mathrm{<span\; class="notranslate">\; Iy</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 6</span>}}\end{array}\right)<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>\end{array}$

Wherein, the Ix represents the gradient map of the grid on the x-axis direction, the Iy represents the gradient map of the grid on the y-axis direction, and the ▽I represents the grid passing through the variability template W(X; p) Transformed gradient map;

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; \&Delta;p</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; *</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

Calculate p5 and p6 when Δp is less than the preset threshold ε;

Obtain the optical flow from the RGB three components of the optical flow information of the grid to obtain the color optical flow information of the grid;

According to the color optical flow information of the grid, calculate the histogram information of the optical flow in 8 directions, the histogram information of the optical flow in 8 directions is the color histogram feature of the optical flow of the grid, the network The color histogram features of the optical flow of the lattice include the optical flow u'b in the horizontal direction and the optical flow _v'b in the vertical direction.

It should be noted that the eight directions are specifically one direction every 45 degrees.

In step S103, for each monitoring scene, cluster the grids in the monitoring scene according to the color histogram information of the optical flow of each grid in the monitoring scene, and obtain the semantic region segmentation of the monitoring scene result.

Specifically, step S103 is implemented as follows:

${\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}{{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

In step S104, the network topology relationship among the cameras in the monitoring network is determined according to the semantic region segmentation result of each monitoring scene.

Specifically, taking two cameras as an example, the two cameras are respectively the first camera and the second camera, and the first and second do not represent the order, but are only used to distinguish the cameras; the first camera acquires the first video stream, the second camera acquires a second video stream, the video stream includes a plurality of frames, each frame is an image, the first video stream includes the first image, and the second video stream includes the second image; in the first image, An image with a moving object passing by is the first monitoring scene, and the moving object includes people, animals or other real objects, and in the second image, an image with a moving object passing by is the second monitoring scene.

${\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; ]</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; c</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}{\mathrm{<span\; class="notranslate">\; arg</span>}\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; \&Psi;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$

表示第一网格的光流的颜色直方图特征和第二网格的光流的颜色直方图特征的关联度，所述

表示第一网格和第二网格的时移，所述Ψ_i,j表示第一摄像机和第二摄像机的拓扑关系估计结果。Wherein, the a _i represents the color histogram feature of the optical flow of the first grid, the first grid is decomposed by the first monitoring scene, and the first monitoring scene is captured by the first camera, and the a _j represents The color histogram feature of the optical flow of the second grid, the second grid is decomposed by the second monitoring scene, the second monitoring scene is taken by the second camera, and the first camera and the second camera are monitoring network Any two cameras in , the c represents the second grid after time τ, the

Indicates the degree of correlation between the color histogram feature of the optical flow of the first grid and the color histogram feature of the optical flow of the second grid, the

represents the time shift of the first grid and the second grid, and the Ψ _i,j represents the topological relationship estimation results of the first camera and the second camera.

It should be noted that when Ψ _i,j is greater than 0.5, it means that the first camera and the second camera are topologically related; in step S104 , it is necessary to calculate the topological relationship estimation result between any two cameras.

Another embodiment of the present invention provides a camera topology estimation result as shown in Figure 2. The camera topology estimation result is as follows:

The embodiment of the present invention selects 7 cameras on the first floor of the administrative building of the Shanghai Institute for Advanced Study, Chinese Academy of Sciences, and selects video streams from 11:00 noon to 1:00 pm on a certain day as samples to calculate the camera topology estimation result.

In the experiment, 7 cameras were deployed on the same floor, among which cameras ① and ③ were deployed at the elevator entrance, and the remaining cameras were deployed at the entrances and exits of the 5 passages. The floor and camera deployment diagram is shown in Figure 3.

The number in the circle in Figure 2 indicates the camera number, which corresponds to the camera number in Figure 3. In the experimental results, there is a solid line connecting the two cameras, indicating that there is a relationship between the monitoring targets of the two cameras, that is, the same target appears in the field of view of the two cameras. It can reflect the target activity trend from the perspective of probability and statistics. No solid line connecting the cameras means that the two cameras have no connection or little connection. For example, there is a strong connection between cameras ①, ⑥, and ⑦. Because the location of camera ⑦ is the main entrance of the entire administrative building, after entering the floor, you must go upstairs through the elevator at camera ① or enter the back end of the first floor through the passage at camera ⑥. canteen. Since the selected time period is lunch time, many people above the second floor need to go to the first floor through the elevator at camera ①, and then enter the cafeteria through the passage at camera ⑥. After lunch, the personnel will return the same way. There is no connection (or very little) between camera ② and camera ③, because the elevator at camera ③ is a freight elevator, which is only used inside the cafeteria, and the two cameras are the back kitchen of the cafeteria, and there is no direct passage.

Another embodiment of the present invention provides a camera network topology relationship estimation system in a monitoring scene. The module structure of the system is shown in FIG. 4, specifically including:

Decomposition unit 41 is used to decompose the monitoring scene in the video stream captured by each camera in the monitoring network into grids;

An acquisition unit 42, configured to acquire, for each monitoring scene, color histogram information of the optical flow of each grid in the monitoring scene;

The clustering unit 43 is configured to, for each monitoring scene, cluster the grids in the monitoring scene according to the color histogram information of the optical flow of each grid in the monitoring scene, to obtain the semantics of the monitoring scene Region segmentation results;

The determining unit 44 is configured to determine the network topology relationship among the cameras in the monitoring network according to the semantic region segmentation result of each monitoring scene.

Optionally, the acquisition unit 42 is specifically configured to:

Define the video stream captured by the camera as I _n (X), where X is the coordinate of the grid in the monitoring scene of the video stream, X=(x; y) ^T , and the x is the horizontal axis of the grid Coordinates, the y is the ordinate of the grid, the T represents the transposition of the matrix, and the n is the number of the video frame included in the video stream;

definition $W(x;p)=\left(\begin{array}{c}(1+p1)\&times;x+p3\&times;\mathrm{the\; y}+p5\\ p2\&times;x+(1+p4)\&times;\mathrm{the\; y}+p6\end{array}\right);---\left(1\right)$

Wherein, W represents the variability template, p=(p1, p2, p3, p4, p5, p6) ^T , the p1, p2, p3, p4 are 0, and the p5, p6 are the optical flow information of the grid;

definition $p=\arg \underset{p}{\min }\underset{x}{\mathrm{\&Sigma;}}{[I\left(W(x;p+\mathrm{\&Delta;p})\right)-T\left(x\right)]}^2;---\left(2\right)$

Among them, Δp represents the difference of p between two iterations, and T(x) represents the decomposed grid of the first frame of the video stream;

Iterate according to (3), (4), and (5) until Δp is less than the preset threshold ε;

$<span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&Delta;x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\begin{array}{c}\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; Ix</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}\\ \mathrm{<span\; class="notranslate">\; Iy</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 6</span>}}\end{array}\right)<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>\end{array}$

Wherein, the Ix represents the gradient map of the grid on the x-axis direction, the Iy represents the gradient map of the grid on the y-axis direction, and the ▽I represents the grid passing through the variability template W(X; p) Transformed gradient map;

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; \&Delta;p</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; *</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

Calculate p5 and p6 when Δp is less than the preset threshold ε;

Obtain the optical flow from the RGB three components of the optical flow information of the grid to obtain the color optical flow information of the grid;

According to the color optical flow information of the grid, calculate the histogram information of the optical flow in 8 directions, the histogram information of the optical flow in 8 directions is the color histogram feature of the optical flow of the grid, the network The color histogram features of the optical flow of the lattice include the optical flow u' _b in the horizontal direction and the optical flow v' _b in the vertical direction.

Optionally, the clustering unit 43 is specifically used for:

${\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

Optionally, the determining unit 44 is specifically configured to:

${\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; ]</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; c</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}{\mathrm{<span\; class="notranslate">\; arg</span>}\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; \&Psi;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$

表示第一网格的光流的颜色直方图特征和第二网格的光流的颜色直方图特征的关联度，所述

表示第一网格和第二网格的时移，所述Ψ_i,j表示第一摄像机和第二摄像机的拓扑关系估计结果。Wherein, the a _i represents the color histogram feature of the optical flow of the first grid, the first grid is decomposed by the first monitoring scene, and the first monitoring scene is captured by the first camera, and the a _j represents The color histogram feature of the optical flow of the second grid, the second grid is decomposed by the second monitoring scene, the second monitoring scene is taken by the second camera, the first camera and the second camera are Any two cameras in the monitoring network, the c represents the second grid after time τ, the

Indicates the degree of correlation between the color histogram feature of the optical flow of the first grid and the color histogram feature of the optical flow of the second grid, the

represents the time shift of the first grid and the second grid, and the Ψ _i,j represents the topological relationship estimation results of the first camera and the second camera.

Optionally, the eight directions are specifically one direction every 45 degrees.

Those of ordinary skill in the art can understand that the various modules included in the above-mentioned embodiments are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be realized; in addition, the specific names of each functional module It is only for the convenience of distinguishing each other, and is not used to limit the protection scope of the present invention.

Those of ordinary skill in the art can also understand that all or part of the steps in the method of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a readable storage medium. Storage media, including ROM/RAM, etc.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention should be included in the protection of the present invention. within range.

Claims (10)

1. a camera network topology relationship estimation method in a surveillance network, is characterized in that, the method comprises: Decompose the monitoring scene in the video stream captured by each camera in the monitoring network into a grid; For each monitoring scene, obtain the color histogram information of the optical flow of each grid in the monitoring scene; For each monitoring scene, cluster the grids in the monitoring scene according to the color histogram information of the optical flow of each grid in the monitoring scene, and obtain the semantic region segmentation result of the monitoring scene; According to the semantic region segmentation results of each monitoring scene, the network topology relationship among the cameras in the monitoring network is determined. 2. The method according to claim 1, wherein said acquiring the color histogram information of the optical flow of each grid in said monitoring scene is specifically: Define the video stream captured by the camera as I _n (X), where X is the coordinate of the grid in the monitoring scene of the video stream, X=(x; y) ^T , and the x is the horizontal axis of the grid Coordinates, the y is the ordinate of the grid, the T represents the transposition of the matrix, and the n is the number of the video frame included in the video stream; definition $W(x;p)=\left(\begin{array}{c}(1+p1)\&times;x+p3\&times;\mathrm{the\; y}+p5\\ p2\&times;x+(1+p4)\&times;\mathrm{the\; y}+p6\end{array}\right);---\left(1\right)$ Wherein, W represents the variability template, p=(p1, p2, p3, p4, p5, p6) ^T , the p1, p2, p3, p4 are 0, and the p5, p6 are the optical flow information of the grid; definition $p=\arg \underset{p}{\min }\underset{x}{\mathrm{\&Sigma;}}{[I\left(W(x;p+\mathrm{\&Delta;p})\right)-T\left(x\right)]}^2;---\left(2\right)$ Among them, Δp represents the difference of p between two iterations, and T(x) represents the decomposed grid of the first frame of the video stream; Iterate according to (3), (4), and (5) until Δp is less than the preset threshold ε; $<span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&Delta;x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\begin{array}{c}\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; Ix</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}\\ \mathrm{<span\; class="notranslate">\; Iy</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 6</span>}}\end{array}\right)<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>\end{array}$ Wherein, the Ix represents the gradient map of the grid on the x-axis direction, the Iy represents the gradient map of the grid on the y-axis direction, and the ▽I represents the grid passing through the variability template W(X; p) Transformed gradient map; $\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$ $\mathrm{<span\; class="notranslate">\; \&Delta;p</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; *</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$ Calculate p5 and p6 when Δp is less than the preset threshold ε; Obtain the optical flow from the RGB three components of the optical flow information of the grid to obtain the color optical flow information of the grid; According to the color optical flow information of the grid, calculate the histogram information of the optical flow in 8 directions, the histogram information of the optical flow in 8 directions is the color histogram feature of the optical flow of the grid, the network The color histogram features of the optical flow of the lattice include the optical flow u' _b in the horizontal direction and the optical flow v' _b in the vertical direction. 3. The method according to claim 2, wherein, for each monitoring scene, the grid in the monitoring scene is processed according to the color histogram information of the optical flow of each grid in the monitoring scene Clustering to obtain the semantic region segmentation result of the monitoring scene is specifically: ${\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}{{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$ ${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$ 4. The method according to claim 3, wherein the network topology relationship between each camera in the monitoring network is determined according to the semantic region segmentation result of each monitoring scene as follows: ${\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; ]</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; c</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$ ${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}{\mathrm{<span\; class="notranslate">\; arg</span>}\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$ ${\mathrm{<span\; class="notranslate">\; \&Psi;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$ Wherein, the a _i represents the color histogram feature of the optical flow of the first grid, the first grid is decomposed by the first monitoring scene, and the first monitoring scene is captured by the first camera, and the a _j represents The color histogram feature of the optical flow of the second grid, the second grid is decomposed by the second monitoring scene, the second monitoring scene is taken by the second camera, and the first camera and the second camera are monitoring network Any two cameras in , the c represents the second grid after time τ, the

Indicates the degree of correlation between the color histogram feature of the optical flow of the first grid and the color histogram feature of the optical flow of the second grid, the

represents the time shift of the first grid and the second grid, and the Ψ _i,j represents the topological relationship estimation results of the first camera and the second camera. 5. The method according to claim 2, wherein the eight directions are specifically one direction every 45 degrees. 6. A camera network topology estimation system in a surveillance network, characterized in that the system comprises: A decomposition unit, configured to decompose the monitoring scene in the video stream captured by each camera in the monitoring network into grids; An acquisition unit, for each monitoring scene, to obtain the color histogram information of the optical flow of each grid in the monitoring scene; A clustering unit, for each monitoring scene, clustering the grids in the monitoring scene according to the color histogram information of the optical flow of each grid in the monitoring scene, to obtain the semantic area of the monitoring scene segmentation result; The determining unit is configured to determine the network topology relationship among the cameras in the monitoring network according to the semantic region segmentation result of each monitoring scene. 7. The system according to claim 6, wherein the acquiring unit is specifically used for: Define the video stream captured by the camera as I _n (X), where X is the coordinate of the grid in the monitoring scene of the video stream, X=(x; y) ^T , and the x is the horizontal axis of the grid Coordinates, the y is the ordinate of the grid, the T represents the transposition of the matrix, and the n is the number of the video frame included in the video stream; definition $W(x;p)=\left(\begin{array}{c}(1+p1)\&times;x+p3\&times;\mathrm{the\; y}+p5\\ p2\&times;x+(1+p4)\&times;\mathrm{the\; y}+p6\end{array}\right);---\left(1\right)$ Wherein, W represents the variability template, p=(p1, p2, p3, p4, p5, p6) ^T , the p1, p2, p3, p4 are 0, and the p5, p6 are the optical flow information of the grid; definition $p=\arg \underset{p}{\min }\underset{x}{\mathrm{\&Sigma;}}{[I\left(W(x;p+\mathrm{\&Delta;p})\right)-T\left(x\right)]}^2;---\left(2\right)$ Among them, Δp represents the difference of p between two iterations, and T(x) represents the decomposed grid of the first frame of the video stream; Iterate according to (3), (4), and (5) until Δp is less than the preset threshold ε; $<span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&Delta;x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\begin{array}{c}\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; Ix</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}\\ \mathrm{<span\; class="notranslate">\; Iy</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; 6</span>}}\end{array}\right)<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>\end{array}$ Wherein, the Ix represents the gradient map of the grid on the x-axis direction, the Iy represents the gradient map of the grid on the y-axis direction, and the ▽I represents the grid passing through the variability template W(X; p) Transformed gradient map; $\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$ $\mathrm{<span\; class="notranslate">\; \&Delta;p</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; *</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}{<span\; class="notranslate">\; [</span><span\; class="notranslate">\; \&dtri;</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ;</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$ Calculate p5 and p6 when Δp is less than the preset threshold ε; Obtain the optical flow from the RGB three components of the optical flow information of the grid to obtain the color optical flow information of the grid; According to the color optical flow information of the grid, calculate the histogram information of the optical flow in 8 directions, the histogram information of the optical flow in 8 directions is the color histogram feature of the optical flow of the grid, the network The color histogram features of the optical flow of the lattice include the optical flow u' _b in the horizontal direction and the optical flow v' _b in the vertical direction. 8. The system according to claim 7, wherein the clustering unit is specifically used for: ${\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}{{\mathrm{<span\; class="notranslate">\; u</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$ ${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; r</span>}}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{{\mathrm{<span\; class="notranslate">\; v</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$ 9. The system according to claim 8, wherein the determining unit is specifically configured to: ${\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; ]</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; c</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$ ${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}{\mathrm{<span\; class="notranslate">\; arg</span>}\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Gamma;</span>}}<span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$ ${\mathrm{<span\; class="notranslate">\; \&Psi;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; \&tau;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}}_{{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$ Wherein, the ai represents the color histogram feature of the optical flow of the first grid, the first grid is decomposed by the first monitoring scene, and the first monitoring scene is captured by the first camera, and the aj represents the second The color histogram feature of the optical flow of the grid, the second grid is decomposed by the second monitoring scene, the second monitoring scene is taken by the second camera, and the first camera and the second camera are monitoring network Any two cameras in , the c represents the second grid after time τ, the

Indicates the degree of correlation between the color histogram feature of the optical flow of the first grid and the color histogram feature of the optical flow of the second grid, the

represents the time shift of the first grid and the second grid, and the Ψ _i,j represents the topological relationship estimation results of the first camera and the second camera. 10. The system according to claim 7, wherein the eight directions are specifically one direction every 45 degrees.

Images