TWI407386B - Method of tracking images automatically - Google Patents

Abstract

The present invention discloses a method of automatically tracking images, which includes: generating related location data by a fixed camera and a PTZ camera; setting an alarm rule in a image analyzer; detecting whether an object in accordance with the alarm rule exists in the frame of the fixed camera by the image analyzer; controlling the PTZ camera to focus on the object by a related module based on the related location data if it objective exists; controlling the PTZ camera to track the object by the related module and the image analyzer; wherein, the related data comprises at least two located points data to define panning, tiling, zooming values of the PTZ camera, and the image analyzer can detect and catch the object continuously to facilitate the PTZ camera to track automatically.

Description

Image automatic tracking monitoring method

The invention relates to an image monitoring method, in particular to a synchronous (zero time difference) image automatic tracking monitoring method.

With the development of technology, the traditional security personnel with the monitoring mode of the general fixed camera can not meet the needs of people for security monitoring. Due to the rapid development of image processing technology and the decline of hardware price, it can be matched with the image processing technology. Surveillance devices have been gradually used, leading to the trend of high-tech security in the field of security monitoring that was originally closed by technology.

In general, the camera can be divided into a fixed camera and a rotary zoom (Pan/Tilt/Zoom, PTZ) camera. The rotary zoom camera has a rotation function and a zoom function, and includes a speed dome camera. A fixed camera can only observe the fixed size of a fixed scene, and cannot rotate and zoom.

A stationary camera monitors all suspicious people, things, and things in their location. However, since there is no rotation or zoom function, even if an intruder is detected in the location where it is located, the fixed camera is still highly likely because the image of the intruder is in a corner or shadow that the fixed camera cannot monitor. Or, because the intruder is too far away from the fixed camera to present a small image in the fixed image, and cannot provide effective information for the user of the fixed camera.

The rotary zoom camera can effectively overcome the above problems faced by fixed cameras. After an intruder is found in the location it is monitoring, the rotary zoom camera can use the rotation function to move the image of the intruder in the rotated zoom image generated by it to the center of the above-mentioned rotated zoom image, and then use the zoom function. Zoom in on the image of the intruder so that the user of the rotating zoom camera can clearly see the intruder's face.

In the current surveillance market, the rotary zoom camera has reached a general trend, and the camera has been moved and changed focal length to achieve the function of monitoring no dead angle. An Intelligent Video System (IVS) or image analyzer can be used to analyze suspicious objects in an image and monitor and track the objects by a rotary zoom camera for automatic monitoring without manual monitoring by the user. When an unexpected event occurs and the location of the event is not the preset position set by the rotary zoom camera, since the location of the incident is not the preset position of the rotating focus camera, it cannot be rotated immediately and focused to the location of the incident, that is, The image analyzer cannot immediately turn the screen to the event occurrence location, which causes the rotary zoom camera and the image analyzer to be effectively used to cause a security monitoring loophole, and the present invention is oyster.

In order to overcome the above disadvantages and difficulties, the present invention provides an instant image automatic monitoring method.

The object of the present invention is to provide an image monitoring method for synchronous positioning and automatic tracking, which uses an fixed camera, a rotating focus camera and an image analyzer to establish an automatic tracking related data, so that the image analyzing device analyzes and detects In the case of a suspicious object, the rotary focus camera can immediately locate and focus on the position of the above suspicious object according to the above-mentioned automatic tracking related data and continuously track the suspicious object, thereby achieving the automatic (or zero time difference) automatic monitoring effect.

In order to achieve the above object, the present invention discloses a method for monitoring automatic image tracking, which comprises: establishing a related positioning data by using a fixed camera and a rotary zoom camera; setting an analysis alarm rule in an image analyzer; Detecting, by the image analyzer, whether there is an object conforming to the analysis alarm rule in the screen of the fixed camera; if yes, controlling, by the associated module, the rotating zoom camera to align the object according to the associated positioning data; Controlling, by the associated module and the image analyzer, the rotating zoom camera to track the object; wherein the associated positioning data includes at least two positioning point data, thereby defining a pan, a tile, and a zoom of the rotary zoom camera. The value (zoom), and the image analyzer can continuously detect and capture the position of the object according to the analysis alarm rule, so as to facilitate the rotation of the zoom camera.

In another aspect, the present invention further provides a method for automatic image monitoring, the method comprising: establishing a camera data by using a device management module; establishing a rotary zoom camera data by using the device management module; Establishing an image analyzer data; establishing an automatic tracking associated data by using an associated module; detecting whether an image analyzer has an alarm by a central management module; if so, the associated module is automatically tracking the association according to the foregoing The data controls a rotary zoom camera to perform automatic tracking; if not, the image monitoring method is ended. In addition, the automatic image monitoring method further includes: using the central management module to detect whether the image of the camera data has a selected object or position; if yes, transferring the captured image of a rotary zoom camera to the image by using the associated module; The object or location selected above.

Another advantage of the present invention is that the above technical features can be set in a plurality of fixed cameras and rotary zoom cameras to achieve a synchronous automatic monitoring function for multiple scenes.

The above is used to clarify the object of the present invention, the technical means for achieving the object, the advantages thereof, and the like. The invention will be apparent to those skilled in the art from the description of the appended claims.

The present invention will be described in terms of the preferred embodiments and aspects of the invention, which are intended to be illustrative and not to limit the scope of the invention. Therefore, the present invention may be widely practiced in other embodiments in addition to the preferred embodiments described in the specification.

As shown in the fifth figure, the image monitoring system of the present invention comprises: a device communication interface 501, an image capturing device 502, a fixed camera 503, a rotary zoom camera 504, a network camera 505, and a network rotation. Zoom camera 506, management system communication interface 601, central management module 602, device management module 603, association module 604, image processing module 605, alarm notification module 606, system database 607 and image analyzer (Intelligent Video System, IVS) 608. In this embodiment, the fixed camera 503 and the rotary zoom camera 504 are both connected to the image capturing device 502 to generate a video stream that can be transmitted through the network, and the image capturing device 502 can include an image capturing card. (Capture Card), Network Video Recorder (NVR) or Digital Video Recorder (DVR), etc. The image capturing device 502, the network camera 505, and the network rotating zoom camera 506 are respectively coupled to the device communication interface 501, so as to transmit the obtained image or video to the central management module 602, wherein the central management module It is coupled to a management system communication interface 601 for wired or wireless transmission with the device communication interface 501, wherein it can be wirelessly transmitted using an ethernet, internet, WiFi wireless network or other communication network. The device management module 603, the association module 604, the image processing module 605, the alarm notification module 606, the system database 607, and the image analyzer 608 are each coupled to the central management module 602, and the central management module 602. The device management module 603, the association module 604, the image processing module 605, and the alarm notification module 606 are all programmable in a processing device of a computer, such as a processor, a digital signal processor (DSP), and a micro Controller, application-specific integrated circuit (ASIC), application specific instruction set processor (ASIP), chip, multi-chip or wafer set, system-on-chip (SoC), multiprocessor system chip (MPSoC) a programmable logic controller (PLC), a graphics processing unit (GPU) or a field programmable gate array (FPGA), and the like, and the system database 607 is used to store camera data and camera related data. It can be a computer readable medium such as a hard disk, a compact disc, a flash memory, a random access memory (RAM), or a read only memory (ROM). The device management module 603 is configured to manage the fixed camera 503, the rotary zoom camera 504, the network camera 505, and the network rotary zoom camera 506, thereby establishing the data of the on-camera camera; and the association module 604 is For establishing camera-related data, and when a specific area of the fixed camera 503 screen is selected or clicked, the center of the screen of the rotary zoom camera 504 is controlled to be immediately transferred to facilitate the shooting of the geographical location selected or clicked on the screen; The processing module converts the video stream obtained by the central management module 602 into a video format that can be played on the screen, such as JPEG, MJPEG, MPEG1, MPEG2, MPEG4, H.263, H.264; Group 606 can alert the user when image analyzer 608 detects a suspicious person or an event occurs.

A method for performing the instant synchronization image automatic monitoring system of the present invention is as follows: first, selecting a fixed camera and selecting a PTZ camera; and then using the fixed camera and the above rotating zoom The camera establishes an associated positioning data, which includes at least two positioning point data, thereby defining a pan, tile, and zoom value of the rotary zoom camera; then, selecting an image analyzer (IVS) and setting An analysis alarm rule is performed on the image analyzer; and then using the image analyzer to detect, according to the analysis alarm rule, whether there is an object that meets (or violates) the analysis alarm rule; if yes, using the association module to locate the data according to the association Controlling the rotary zoom camera to align the object; finally, controlling the rotating zoom camera to track the object by the associated module and the image analyzer. Therefore, the rotary zoom camera disclosed in the present invention can immediately transfer the screen to the suspicious object measured by the image analyzer according to the associated positioning data, and continuously track the object according to the detection result of the image analyzer, thereby achieving instant detection. Automatic image monitoring.

Please refer to the first figure, which is a flowchart of a preferred embodiment of the present invention, and discloses a method for automatic image monitoring, the steps of which are as follows: In step 101, an image automatic monitoring system is activated, The image automatic monitoring system comprises at least one fixed camera and at least one rotary zoom camera (Pan/Tilt/Zoom, PTZ), wherein the rotary zoom camera is preferably a speed dome camera; then, in step 102 In this system, a central management module detects whether a camera data exists in the system database, and the camera data refers to image data generated by a fixed camera. Specifically, the camera data system For video streaming, it can be compressed into JPEG, MJPEG, MPEG1, MPEG2, MPEG4, H.263, H.264, etc.; however, if there is no camera data in the system database, then enter Step 103: Establish a camera data by using a device management module, that is, obtain image data of the fixed camera by using the device management module, and store the image data in the system database. Then, if the camera data already exists in the system database, proceed to step 104, and the central management module detects whether at least one rotating zoom camera data exists in the system database, and the rotating zoom camera data is rotated. The image data generated by the zoom camera is specifically a video stream; however, if there is no rotating zoom camera data in the system database, the process proceeds to step 105, and the device management module is used to establish Rotating the zoom camera data, that is, using the device management module to obtain the image data of the rotary zoom camera, and storing it in the system database; however, if there is the rotating zoom camera data, proceeding to step 106, centrally managed The module detects whether at least one image analyzer data exists in the system database. Specifically, the image analyzer data refers to a process of image processing for analyzing the captured image, which can be based on Designed by the user's needs, since this image calculation process is not the focus of the present invention, This is not described in detail; however, if there is no image analyzer data in the system database, proceed to step 107, and use the device management module to create an image analyzer data, that is, obtain an image of the image analyzer. The calculation process or the image algorithm; then, proceeding to step 108, the central management module detects whether there is automatic tracking related data in the system database, and the automatic tracking related data includes the associated positioning data and the analysis alarm rule, wherein the association The positioning data includes at least two positioning point data, which is used to define the pan, tile and zoom values required for the shooting image of the rotary zoom camera to be positioned at various positions in the fixed camera. The rotary zoom camera can be quickly positioned, so that the lens of the rotary zoom camera can capture objects detected by the image analyzer (for example, but not limited to people, vehicles, etc.), and the detailed positioning method will be The instructions are further described, and the analysis alarm rules are set by the user to determine the object to be monitored or If the automatic tracking related data does not exist in the system database, the process proceeds to step 109, and the associated module establishes automatic tracking related data. The detailed execution process of this step 109 will be further described in this specification. Then, in step 110, the association control is performed by using the associated module, and the associated control refers to when the image analyzer detects an object that meets the analysis alarm rule, or the user selects the fixed camera. When an object or position is used, the rotary zoom camera can react immediately, moving the zoom zoom camera lens so that the image captured by the lens contains the above object or position, preferably in the central area of the captured image, that is, the screen image Center, and can focus on the above object; then, in step 111, the central management module detects whether the image analyzer has an alarm, that is, whether there is a compliance (or violation) analysis alarm rule in the screen of the fixed camera. The object; if yes, proceed to step 112 to perform automatic tracking, and the detailed process will be described in this specification In addition, if not, proceed to step 115, the user decides whether to end the monitoring program, or proceeds to step 113, and the central management module detects whether the fixed camera screen is selected or clicked, more specifically In one aspect, the present invention includes a user interface that displays a captured image or image of a stationary camera and displays a cursor on the display for the user to select or click on the object or location to be viewed; if selected or clicked, Then, in step 114, the associated module performs the associated control to move the screen of the rotary zoom camera to the selected or clicked object or position, and can focus on the object or position; if not selected or clicked, enter In step 115, the user decides whether to end the monitoring program. Finally, if the end monitoring program is selected, then in step 116, the image monitoring method ends.

Please refer to the second figure. This figure describes in detail the method for establishing automatic tracking related data in step 109, which can be roughly divided into two parts: establishing associated positioning data (steps 201-209) and setting analysis alarm rules (steps 210-215). The step of establishing the associated positioning data is as follows: In step 201, a fixed camera is selected; then, in step 202, a rotating zoom camera is selected, which is a rotating zoom associated with the fixed camera described above. The camera, that is, the rotary zoom camera must be able to capture the field of view of the selected fixed camera; then, in step 203, the associated module detects whether there are two positioning points in the system database; If the data of the two positioning points already exists, the process proceeds to step 204, where the name of the associated positioning data is defined. Specifically, the user can input the text or number through the user interface to make the name of the associated data; then, proceed to step 205. The storage associated positioning data is stored in the system database; however, returning to step 203, if there is no data of the two positioning points in the system database, the process proceeds to step 206; in step 206, the fixed camera is used to set a The positioning point, in particular, can be selected or clicked by the user in an image taken by the fixed camera or Position, which is defined as an anchor point; then, in step 207, the lens of the rotary zoom camera is controlled to cause the selected object or position to be presented in the image or screen center captured by the rotary zoom camera. Positioning program. Specifically, the user can control the rotating zoom camera to rotate left and right (pan) and up and down (tile) to align and fix the center of the screen to the positioning point; then, in step 208, adjust the Rotate the zoom camera to match the size of the object. Specifically, use the zoom function of the rotary zoom camera to zoom the zoom camera and focus on the anchor point to make the object at the anchor point Or the position of the rotating zoom camera occupies a certain proportion of the area and is clearly visible, so as to achieve the effect of monitoring.

Then, in step 209, the above-mentioned positioning point data is stored in the system database, specifically, the left and right rotation (pan), the up and down tilt (Tile) and the zoom size of the zoom camera when the positioning point is rotated. The zoon) values are defined as (X ₁ , Y ₁ , Z ₁ ), respectively, and are stored as an anchor point data. Then, returning to step 203, steps 206-209 are repeated to establish data of another positioning point (X ₂ , Y ₂ , Z ₂ ), and according to the data of the two positioning points, the association can be defined in step 204. The name of the location data is stored and stored in step 205. The associated positioning data utilizes the above two positioning points to calculate the (X, Y, Z) required to align the rotary zoom camera to other points in the fixed camera screen, that is, the shooting image of the rotary zoom camera. The pan, tile, and zoom required to move to each of the above points are calculated in the following manner: Define a plane coordinate (x, y) in the fixed camera's picture, assuming The first positioning point is (x ₁ , y ₁ ) in the picture of the fixed camera, and the second positioning point is (x ₂ , y ₂ ), and the rotary zoom camera is rotated from the origin to the first positioning point. The pan value is X ₁ , and the pan value to the second anchor point is X ₂ , then an absolute rotation parameter Base_P=((X ₁ *(x ₂ /x ₁ ))−X ₂ )/(x ₂ / x ₁ -1), and thus the absolute rotation parameter defines a relative rotation parameter Com_P=(X ₁ -Base_P)/x ₁ , and the absolute rotation parameter Base_P and the relative rotation parameter Com_P can be calculated in the screen of the fixed camera The pan value X of other positions to be observed is as follows:

X=Base_P+x*Com_P

Where x is the abscissa value to be observed in the fixed camera screen, and the pan value X at which the rotary zoom camera rotates to this position can be obtained by the above relationship. Similarly, assuming that the rotary zoom camera is tilted from the origin to the tile value of the first anchor point Y ₁ and the tile value of the second anchor point is Y ₂ , an absolute tilt parameter Base_T=((Y ₁ *) can be defined. (y ₂ /y ₁ ))-Y ₂ )/(y ₂ /y ₁ -1), and thus the absolute tilt parameter defines a relative tilt parameter Com_T=(Y ₁ -Base_T)/y ₁ , and is absolutely tilted The parameter Base_T and the relative tilt parameter Com_T can calculate the Y value of other positions to be observed in the screen of the fixed camera, and the relationship is as follows:

Y=Base_T+y*Com_T

Where y is the ordinate value to be observed in the fixed camera screen, and the above-mentioned relationship can be used to obtain the tile value Y at which the rotary zoom camera rotates to this position. Finally, assuming that the zoom value of the rotary zoom camera at the first anchor point is Z ₁ and the zoom value of the second anchor point is Z ₂ , an absolute zoom parameter Base_Z=((Z ₁ *(y) can be defined. ₂ / y ₁ )) - Z ₂ ) / (y ₂ / y ₁ -1), and thus the absolute scaling parameter defines a relative scaling parameter Com_Z = (Z ₁ -Base_Z) / y ₁ , and the absolute scaling parameter Base_Z And the relative scaling parameter Com_Z can calculate the Z value of other positions to be observed in the screen of the fixed camera, and the relationship is as follows:

Z=Base_Z+y*Com_Z

Where y is the ordinate value to be observed in the fixed camera screen, and the zoom value Z of the rotary zoom camera rotated to this position can be obtained by using the above relationship. According to this, as long as the two positioning points are defined by the fixed camera and the rotary zoom camera, the parameter values of the above relational expression can be obtained, and the respective positions (x, y) in the screen of the fixed camera can be transmitted through the above relational expression. Converted to (X, Y, Z) values, that is, the pan, tile, and zoom values required for the rotary zoom camera to position from the origin to each position, thereby enabling image analysis Immediately when any point (x, y) in the fixed camera is detected to trigger an alarm in accordance with the analysis alarm rule, or when any point (x, y) in the fixed camera screen is selected or clicked, the associated module can be immediately Convert (x, y) to (X, Y, Z) value and control the position of the rotary zoom camera to trigger the alarm or the position of the selected click according to the (X, Y, Z) value, thereby achieving instant, zero Time difference monitoring.

After the establishment of the associated positioning data is completed, the step of setting the analysis alarm rule may be entered. The detailed process is as follows: In step 210, it is detected whether the analysis alarm rule exists in the image analyzer; if yes, proceed to step 215. The associated positioning data and the analysis alarm rule are stored as automatic tracking related data; if there is no analysis alarm rule in the image analyzer, the process proceeds to step 211, where the user selects the detection rule type, specifically The detection rule refers to how to determine the analysis range of the image analyzer, which can be roughly divided into three types, namely, a half line type, a double line type, and a range type, wherein the half line type means that the user only needs to be in a fixed camera. One line is drawn in the shooting picture, and one side of the line is the analysis range of the image analyzer; the double half line type means that the user needs to draw two lines in the shooting picture of the fixed camera, by the two lines The divided partial blocks are the analysis range of the image analyzer; and the regional range type is selected by the user in the shooting picture of the fixed camera. A specific range, the range block in or outside the circle is the analytical range of the image analyzer. Then, in step 212, the selected image object type is selected. Specifically, the step is used to determine the type of the object to be analyzed in the analysis range of the image analyzer, which may include, but is not limited to, a traffic vehicle (such as a car, a motorcycle, a bicycle, etc.), a person, an animal, or any moving object, etc.; then, in step 213, the user draws a detection rule on the display screen of the fixed camera. Specifically, if the detection rule type selected in step 211 is a half line type, the user can draw any line on the display screen; if the selected detection rule type is a double half line type, then use The user can draw any two lines on the display screen; if the selected detection rule type is the area range type, the user can draw at least one closed figure on the display screen, such as a circle, a rectangle, or a triangle. Wait. Next, in step 214, the detection rule and the detected image object type are stored as an analysis alarm rule. Finally, in step 215, the associated location data and the analysis alert rule are stored as automatic tracking associated data.

Referring to the third figure, this figure describes the process of automatic tracking in step 112, and the steps are as follows: First, in step 301, whether there is automatic tracking related data in the search system database; then, in step 302, The image analyzer is used to capture an object that triggers an alarm in a fixed camera screen. Specifically, the image analyzer can detect and capture (or violate) the captured image of the fixed camera according to the analysis alarm rule set by the user. The object of the alarm rule is analyzed. Then, in step 303, the associated module is used to calculate the required distance of the shooting screen of the rotary zoom camera to move to the object. Specifically, it is assumed that the object triggering the alarm is located at point A, and the coordinates are (x ₃ , y ₃ ), the association module can convert (X ₃ , Y ₃ , Z ₃ ) according to the above relationship defined by the two positioning points, that is, the rotary zoom camera is positioned from the origin ( x ₃ , y ₃ ) required pan, tile, and zoom values, assuming that the center of the screen of the current rotating zoom camera corresponds to point B, and the coordinates are (x ₄ , y ₄ ). Correlation mode The group can convert it to (X ₄ , Y ₄ , Z ₄ ) according to the above relationship, that is, the rotation (pan) and tilt required by the rotary zoom camera to locate (x ₄ , y ₄ ) from the origin. Tile) and zoom value, and then use the correlation module to calculate the difference between (X ₃ , Y ₃ , Z ₃ ) and (X ₄ , Y ₄ , Z ₄ ) (△X ₁ , ΔY ₁ , △Z ₁ ), this is the required distance to move the screen of the rotary zoom camera from the current position (point B) to the object position (point A); then, in step 304, according to the required distance (ΔX ₁ , △ Y ₁ , ΔZ ₁ ), the lens of the rotary zoom camera is controlled, and the display screen is aligned with the object position. Due to the associated positioning data established by the present invention, each position (x, y) in the screen of the fixed camera can be converted into an (X, Y, Z) value, that is, the rotary zoom camera is positioned from the origin to each position. The required pan, tile, and zoom values, so when the image analyzer detects an object that triggers an alarm, the rotary zoom camera can immediately move its image to the associated positioning data. The position of the object is further improved by the automatic monitoring of the zero time difference; then, in step 305, the image analyzer detects whether the object is still present in the shooting picture of the fixed camera; if not, the process proceeds to step 306. The lens of the rotary zoom camera is controlled by the associated module to be moved back to the preset position; however, if the object is still in the screen of the fixed camera, then returning to step 302, the image analyzer detects and captures A position below the object, and then through steps 303 and 304, calculate and control the rotary zoom camera to be aligned with a position below the object, and finally proceed to step 305 to detect whether the object is still fixed. Camera's screen, and if so, repeat steps 302-304 continue to monitor this object. Therefore, as long as the object is still in the shooting range of the fixed camera, the image analyzer can detect and capture it regardless of its position, so the rotary zoom camera can immediately monitor and continuously track the object to achieve synchronization. The image is automatically tracked.

Referring to the fourth figure, this figure describes in detail the method of moving the screen of the rotary zoom camera to the selected object or position in step 114, the steps of which are as follows: First, in step 401, the association in the system database is searched. Positioning the data; then, proceeding to step 402, using the associated module to capture the selected or clicked object or position in the fixed camera screen, specifically, when the user moves the cursor to the display screen of the fixed camera screen When an object or location is clicked, the associated module can capture the object or location; then, in step 403, the associated module is used to calculate the required distance of the rotating zoom camera to move the object or position, specifically In other words, if the clicked position is C point and coordinates (x ₅ , y ₅ ), the associated module can convert it to (X ₅ , Y ₅ , Z ₅ ) according to the above relationship defined by the two anchor points. That is, the rotary zoom camera is positioned from the origin to the pan, tile, and zoom values required for (x ₅ , y ₅ ), assuming that the center of the screen of the currently rotating zoom camera still corresponds to point B. , Denoted (x _4, y _4), associated with the module according to the above relation, it is converted to _{(X 4, Y 4, Z} 4), i.e. of formula rotate the zoom camera to locate the origin (x _4, y ₄ ) The required pan, tile and zoom values, then use the correlation module to calculate the difference between (X ₅ , Y ₅ , Z ₅ ) and (X ₄ , Y ₄ , Z ₄ ) (ΔX ₂ , ΔY ₂ , ΔZ ₂ ), which is the required distance to move the screen of the rotary zoom camera from the current position (point B) to the selected object or position (point C); In step 404, according to the required distance (ΔX ₂ , ΔY ₂ , ΔZ ₂ ), the screen of the rotary zoom camera is controlled to be aligned with the object or position, thereby achieving the effect of real-time monitoring.

The above description is a preferred embodiment of the invention. Those skilled in the art should be able to understand the invention and not to limit the scope of the patent claims claimed herein. The scope of patent protection is subject to the scope of the patent application and its equivalent fields. Any modification or refinement made by those skilled in the art without departing from the spirit or scope of the present invention is equivalent to the equivalent change or design made in the spirit of the present disclosure, and should be included in the following patent application scope. Inside.

101-116. . . step

201-215. . . step

301-306. . . step

401-404. . . step

501. . . Device communication interface

502. . . Image capture device

503. . . Fixed camera

504. . . Rotary zoom camera

505. . . Webcam

506. . . Network rotary zoom camera

601. . . Management system communication interface

602. . . Central management module

603. . . Device management module

604. . . Association module

605. . . Image processing module

606. . . Alarm notification module

607. . . System database

608. . . Image analyzer

The first figure is a flow chart of a preferred embodiment of the present invention;

The second figure is a flow chart for establishing automatic tracking related data in step 109;

The third figure is a flow chart of automatic tracking in step 112;

The fourth figure is a flow chart of moving the screen of the rotary zoom camera to the selected object or position in step 114;

The fifth figure is an image monitoring system used in the image monitoring method of the present invention.

101-116. . . step

Claims (11)

A method for monitoring automatic image tracking, comprising the steps of: establishing a related positioning data by using a fixed camera and a rotating zoom camera; setting an analysis alarm rule to an image analyzer; and detecting whether the image analyzer exists by using the image analyzer Aligning one of the analysis alarm rules with the object of the fixed camera; if yes, controlling the rotary zoom camera to align the object according to the associated positioning data; and the associated module and the image The analyzer controls the rotary zoom camera to automatically track the object; wherein the associated positioning data includes at least two positioning point data, thereby defining a pan, a tile, a zoom value of the rotary zoom camera, and The image analyzer continuously detects the object according to the analysis alarm rule to facilitate automatic tracking of the rotary zoom camera. The method for monitoring automatic image tracking according to claim 1, wherein the step of establishing the positioning point data comprises: setting an positioning point by using the fixed camera; controlling the rotating zoom camera to rotate and tilting, so that the rotating zoom camera The captured image includes the positioning point; the image of the rotating zoom camera is scaled to conform to the object size; the rotation, tilt and zoom values of the rotary zoom camera are defined as the positioning point data. The method for monitoring automatic image tracking according to claim 1, wherein the step of setting the analysis alarm rule comprises: selecting a detection rule type; selecting a detection image object type; depicting in the screen of the fixed camera The detection rule; stored as the analysis alarm rule. The method for automatically tracking an image according to claim 1, wherein the step of automatically tracking the object comprises: capturing the object by the image analyzer; and calculating, by using the association module, the screen of the rotary zoom camera is transferred to The distance required by the object is controlled by the association module to transfer the image of the rotating zoom camera to the object; and the image analyzer is used to detect whether the object exists in the captured image of the fixed camera; if so, the associated mode The group controls the rotary zoom camera to track the object. A method for monitoring automatic image tracking includes the steps of: establishing a camera data by a device management module; establishing a rotary zoom camera data by using the device management module; and establishing an image analyzer data by using the device management module; Establishing an automatic tracking associated data by using an associated module; detecting whether an image analyzer has an alarm by a central management module; if so, controlling, by the associated module, a rotating zoom camera according to the automatic tracking associated data To perform automatic tracking. The method for monitoring automatic image tracking according to claim 5, further comprising: detecting, by the central management module, whether an object or a location is selected in the camera data; if yes, using the associated module to rotate the zoom The image captured by the camera is transferred to the selected object or location. The method for monitoring automatic image tracking according to claim 5, wherein the step of establishing an automatic tracking related data comprises: selecting a fixed camera; selecting the rotary zoom camera; and determining, by using the associated module, the automatic tracking related data Whether to include two positioning point data; if not, using the fixed camera and the rotating zoom camera to perform an associated positioning step; if yes, storing the two positioning point data as an associated positioning data; selecting an image analyzer; utilizing The central management module determines whether an analysis alarm rule has been set in the image analyzer; if so, the associated location data and the analysis alarm rule are stored as an automatic tracking associated data; if not, an analysis alarm rule is set . The method for monitoring automatic image tracking according to claim 7, wherein the associated positioning step comprises: setting an positioning point by using the fixed camera; controlling the positioning of the rotating zoom camera to the positioning point; and zooming the rotating zoom camera The image is to match the size of the object. The method for monitoring automatic image tracking according to claim 7, wherein the step of setting the analysis alarm rule comprises: selecting a detection rule type; selecting a detection image object type; in the screen of the fixed camera; Describe rules; save as the analysis alert rule. The method for automatically tracking an image according to claim 5, wherein the step of automatically tracking comprises: capturing, by the image detector, an object that triggers the alarm; and using the association module to calculate a captured image of the rotary zoom camera; Transferring the required distance to the object; using the association module to control the captured image of the rotary zoom camera to be transferred to the object; using the associated module to control the captured image of the rotary zoom camera and tracking the object according to the automatic tracking associated data . The method for monitoring automatic image tracking according to claim 6, wherein the step of transferring a screen of the rotary zoom camera to the selected object or position comprises: using the associated module to retrieve the selected object or Positioning; calculating, by using the association module, a required distance that the captured image of the rotary zoom camera moves to the selected object or position; and using the associated module to control the captured image of the rotary zoom camera to move to the selected object Or location.