CN111160155A - Method and device for detecting stagnant water - Google Patents
Method and device for detecting stagnant water Download PDFInfo
- Publication number
- CN111160155A CN111160155A CN201911300910.0A CN201911300910A CN111160155A CN 111160155 A CN111160155 A CN 111160155A CN 201911300910 A CN201911300910 A CN 201911300910A CN 111160155 A CN111160155 A CN 111160155A
- Authority
- CN
- China
- Prior art keywords
- detection
- stagnant water
- water
- area
- frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
本发明涉及人工智能技术领域，公开了一种积水检测方法及装置。该方法包括：获取待检测视频图像帧和积水检测模型；所述待检测视频图像帧通过所述积水检测模型进行积水检测，获取检测结果；如果所述检测结果为存在积水，获取目标图像的积水检测区域位置属性信息。采用本发明技术方案不但可以精确的确定积水区域，且可靠性和鲁棒性较高。
The invention relates to the technical field of artificial intelligence, and discloses a method and a device for detecting accumulated water. The method includes: acquiring a video image frame to be detected and a water accumulation detection model; performing water accumulation detection on the video image frame to be detected through the water accumulation detection model, and acquiring a detection result; if the detection result is that there is accumulation of water, acquiring The location attribute information of the water accumulation detection area of the target image. By adopting the technical scheme of the present invention, not only the water accumulation area can be accurately determined, but also the reliability and robustness are high.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a method and a device for detecting accumulated water.
Background
At present, urban road ponding seriously influences the traveling of people, and particularly after heavy rainfall, a sewage discharge system is not in place, so that the formation of large-area urban ponding and even urban waterlogging can be accelerated. In addition, the water pipe bursts, and the circumstances such as groundwater spills over also often takes place in the city, if can not in time handle, not only extravagant urban water resource, still to urban environment, traffic etc. bring unnecessary trouble, consequently, before forming large tracts of land ponding even waterlogging, can be timely accurate detect out road ponding and early warning can effectively reduce the loss. The traditional accumulated water detection mode is usually to detect by a hardware mode, such as a polarization measurement method for detecting the accumulated water on the road surface. A polaroid with a horizontal projection direction is fixed in front of a CCD image sensor of a road condition camera, a TN type liquid crystal is placed in front of the polaroid, a control voltage circuit is arranged on the TN type liquid crystal, and the polarization characteristic of light is utilized, and the water accumulation condition of a road surface is detected by the polarized light of the reflected light of the road surface. In the prior art, two cameras are used for respectively obtaining a first road image picture under a horizontal polarizing film and a second road image picture under a vertical polarizing film, whether the brightness difference value of the two groups of pictures is larger than a preset brightness range is judged on the premise that the two cameras do not exceed a preset value, and if yes, the road surface in the shooting range is judged to have water accumulation.
In the implementation process of the prior art, the inventor finds that the prior art has at least the following technical problems:
in the prior art, the method for detecting the accumulated water on the road surface is complex to realize, high in cost and low in stability, and influences the accumulated water detection result, so that the reliability and robustness of the detection result of the system are low.
Disclosure of Invention
The invention aims to provide a method and a device for detecting accumulated water, which aim to overcome the defects that the method for detecting the accumulated water on the road surface in the prior art is complex in realization, high in cost, low in stability, and low in reliability and robustness of the detection result of the system due to the influence on the accumulated water detection result.
In order to solve the technical problem, an embodiment of the present invention provides a method for detecting standing water, including:
acquiring a video image frame to be detected and a water accumulation detection model;
carrying out water accumulation detection on the video image frame to be detected through the water accumulation detection model to obtain a detection result;
and if the detection result is that the accumulated water exists, acquiring the position attribute information of the accumulated water detection area of the target image.
In order to solve the above technical problem, an embodiment of the present invention further provides a device for detecting standing water, including:
the information acquisition unit is used for acquiring a video image frame to be detected and an accumulated water detection model;
the detection unit is used for carrying out water accumulation detection on the video image frame to be detected through the water accumulation detection model to obtain a detection result;
and the information output unit is used for acquiring the attribute information of the position of the ponding detection area of the target image if the detection result shows that the ponding exists.
The invention provides a method and a device for detecting accumulated water, which are characterized in that a video image frame to be detected and an accumulated water detection model are obtained; carrying out water accumulation detection on the video image frame to be detected through the water accumulation detection model to obtain a detection result; and if the detection result is that the accumulated water exists, acquiring the position attribute information of the accumulated water detection area of the target image. By adopting the technical scheme of the invention, the ponding area can be accurately determined, and the reliability and the robustness are higher.
Drawings
Fig. 1 is a flowchart of a method for detecting standing water according to an embodiment of the present invention;
FIG. 2 is a schematic structural diagram of a water accumulation detection device according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a target detection update architecture centret adopted in the water accumulation detection method according to the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it will be appreciated by those of ordinary skill in the art that numerous technical details are set forth in order to provide a better understanding of the present application in various embodiments of the present invention. However, the technical solutions claimed in the claims of the present application can be implemented without these technical details and with various changes and modifications based on the following embodiments.
A first embodiment of the invention relates to a method of detecting standing water. The specific flow is shown in figure 1.
The method comprises the following steps:
101: acquiring a video image frame to be detected and a water accumulation detection model;
102: carrying out water accumulation detection on the video image frame to be detected through the water accumulation detection model to obtain a detection result;
103: and if the detection result is that the accumulated water exists, acquiring the position attribute information of the accumulated water detection area of the target image.
It should be noted that the location attribute information of the ponding detection area includes: detecting frames of the water accumulation areas and coordinates and categories of the position points; the method further comprises the following steps:
acquiring a detection frame with a water accumulation area and a labeled sample set of position point coordinates;
performing data enhancement processing on the labeled sample set to obtain an expansion labeled sample set;
and training the ponding detection model according to the expansion labeling sample set.
It should be further noted that, the method further includes:
acquiring the accumulated water detection model;
and converting the single-precision accumulated water detection model into a half-precision model.
It should be further noted that, the step of performing data enhancement processing on the labeled sample set to obtain an expanded labeled sample set includes:
acquiring a water accumulation area picture set in a marked sample set according to the detection frame with the water accumulation area and the position point coordinates;
performing local data enhancement processing on the ponding region picture set to obtain an expansion ponding region picture set;
and acquiring a capacity expansion labeling sample set according to the capacity expansion waterlogged area picture set. The specific implementation of the step is as follows: the image set of the expanded ponding area after the local data enhancement is also required to be randomly attached to a road for simulating real ponding and generating the label information of the ponding area on the corresponding composite road. That is to say, the expansion ponding region picture set is combined with the random road to form an expansion labeling sample set.
It should be further noted that the detection result includes: the positions and the scores of the detection frames of the ponding detection model and the positions and the scores of the masks of the ponding detection model; the method further comprises the following steps:
presetting a ponding detection threshold;
judging the mask score of the detection frame score output by the ponding detection model and the ponding detection threshold;
and if the detection frame score and the mask score exceed the ponding detection threshold, determining that the detection frame is an effective detection frame and an effective mask, and simultaneously outputting a detection frame coordinate and a mask coordinate, namely a ponding area. In this step, the detection result includes not only the detection frame score and position (the detection frame is the detection frame for accumulated water), but also the coordinate and score of the accumulated water area, and if the score of the detection frame and the score of the accumulated water area are greater than the set accumulated water detection threshold, the detection frame and the accumulated water area are output.
A second embodiment of the invention relates to a water accumulation detecting device. As shown in fig. 2. The device includes:
an information obtaining unit 201, configured to obtain a video image frame to be detected and a water accumulation detection model;
the detection unit 202 is configured to perform water accumulation detection on the video image frame to be detected through the water accumulation detection model to obtain a detection result;
and the information output unit 203 is used for acquiring the position attribute information of the ponding detection area of the target image if the detection result is that ponding exists.
It should be noted that the location attribute information of the ponding detection area includes: detecting frames of the water accumulation areas and coordinates and categories of the position points; the device also includes:
the sample information acquisition unit is used for acquiring a detection frame with a ponding area and a labeled sample set of position point coordinates;
the data enhancement unit is used for carrying out data enhancement processing on the labeled sample set to obtain an expansion labeled sample set;
and the model training unit is used for training the accumulated water detection model according to the expansion labeling sample set.
It should be further noted that the apparatus further includes:
and the model optimization unit is used for converting the single-precision accumulated water detection model into a half-precision model.
It should be further noted that the data enhancement unit is further configured to obtain a water accumulation region picture set in the labeled sample set according to the detection frame with the water accumulation region and the position point coordinates; performing local data enhancement processing on the ponding region picture set to obtain an expansion ponding region picture set; and acquiring a capacity expansion labeling sample set according to the capacity expansion waterlogged area picture set.
It should be further noted that the detection result includes: the positions and the scores of the detection frames of the ponding detection model and the positions and the scores of the masks of the ponding detection model; the device also includes:
the device comprises a presetting unit, a detection unit and a control unit, wherein the presetting unit is used for presetting a ponding detection threshold;
the judging unit is used for judging the mask score of the detection frame score output by the ponding detection model and the ponding detection threshold;
and the output unit is used for determining the detection frame as an effective detection frame and an effective mask if the detection frame score and the mask score exceed the ponding detection threshold, and simultaneously outputting a detection frame coordinate and a mask coordinate, namely a ponding area.
It should be understood that this embodiment is an example of the apparatus corresponding to the first embodiment, and may be implemented in cooperation with the first embodiment. The related technical details mentioned in the first embodiment are still valid in this embodiment, and are not described herein again in order to reduce repetition. Accordingly, the related-art details mentioned in the present embodiment can also be applied to the first embodiment.
It should be noted that each module referred to in this embodiment is a logical module, and in practical applications, one logical unit may be one physical unit, may be a part of one physical unit, and may be implemented by a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, elements that are not so closely related to solving the technical problems proposed by the present invention are not introduced in the present embodiment, but this does not indicate that other elements are not present in the present embodiment.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
For convenience of description, the above devices are described separately in terms of functional division into various units/modules. Of course, the functionality of the units/modules may be implemented in one or more software and/or hardware implementations of the invention.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
Based on the embodiment, the ponding detection principle is combined, the convolutional neural network is utilized to train the ponding detection model, the detection frame of the ponding detection area and the coordinates of the position point are output, the training data of the model can be effectively increased by further adopting a virtual data enhancement method, the detection rate is further effectively improved, the TensorRT can be adopted to convert the single-precision model into the semi-precision model, the scale of the ponding detection model is optimized, and the detection efficiency is improved. In the implementation process of the technical scheme, a camera is used for collecting a video (namely a video image frame to be detected) of a target area, the video is transmitted back to a terminal processor through a network cable, the terminal processor detects a video single frame (namely the video image frame to be detected) by using a water accumulation detection model, if the confidence degrees of a detection frame and a mask exceed a set value (namely a water accumulation detection threshold value), the position of the detection frame is a road water accumulation detection frame, and the coordinate of the mask is an area where road water accumulation is located. The method can be used in any scene where a camera is present, including but not limited to highways, street roads, commercial districts, parks, attractions, etc., with expandability and portability! By means of repeated training and a TensorRT optimization model method, false detection rate of ponding detection is effectively reduced, detection efficiency is improved, and the method has high reliability and stability. The technical scheme of the invention is concretely realized as follows:
firstly, training a ponding detection model; before the ponding detection model needs to be trained, a training sample set needs to be collected; the training sample set is a marked sample set with a detection frame of a ponding area and position point coordinates; the training method of the ponding detection model comprises the following steps: acquiring a detection frame with a water accumulation area and a labeled sample set of position point coordinates; performing data enhancement processing on the labeled sample set to obtain an expansion labeled sample set; and training the ponding detection model according to the expansion labeling sample set.
The data enhancement processing adopts a virtual enhancement data method to increase training data, and adopts a convolutional neural network to train a ponding detection model. Then, the ponding detection model can be optimized by using TensorRT.
Based on the trained ponding detection model, the ponding detection method disclosed by the invention is specifically realized by the following steps:
step 1: the method comprises the steps that a terminal processor obtains a video image frame to be detected and a water accumulation detection model;
step 2: and (2) transmitting the video collected by the camera back to a terminal processor (namely the video image frame to be detected obtained in the step (1)) through a network cable, and carrying out ponding detection on the water part of the road area in the video single-frame image (namely the video image frame to be detected) by the terminal processor, and outputting a detection frame of a ponding area and the coordinates and categories of the position points.
And step 3: and automatically storing the detection result, the time and the place and other information and transmitting the detection result to a terminal display screen.
It should be noted that, in step 2, the target detection update architecture centret is used as a detection network, and the process of training the ponding detection model is as follows:
a. collecting 6000 pictures with surface water in different scenes, randomly selecting 4000 pictures as a training set, and 2000 pictures as a verification set;
b. marking the region with accumulated water in the picture in a way of simultaneously marking a detection frame and a coordinate point of the accumulated water region;
c. in order to adapt to different scenes, training data are added by a virtual data enhancement method besides data enhancement of the training pictures; the method comprises the following specific steps:
copying the ponding area in the training set by using a polygonal frame, obtaining a clean ponding area by using a background removing algorithm, performing local data enhancement such as stretching, compression, brightness, contrast, color adjustment and the like on the ponding area, randomly pasting the adjusted ponding area on other pictures with road surfaces for simulating a ponding target, and adding a label for simulating the ponding target in an original label file. The original picture was randomly cropped to 500x500 centered on the water, with the picture that contains no targets (i.e., no water) as a negative sample. The number of original training sets was expanded to 30000 by the above procedure.
The ponding detection model is an improved CenterNet model; the improved CenterNet model is additionally provided with a mask prediction module on the basis of the original CenterNet model, wherein the mask is the position point coordinates and the category of the water accumulation area; as shown in fig. 3, an improved centret model proposed for the technical solution of the present invention includes: the device comprises a feature extraction network module, a detection frame prediction module, a mask prediction module and a coupling output module; the feature extraction network module is a typical hourglass network, the detection frame prediction module is a detection head part of an original CenterNet, the mask prediction module is a mask prediction head part added in the method, and the coupling output module is added in the technical scheme of the invention in order to output the position point coordinates of the detection frame and the water accumulation area at the same time. And the video image frame to be detected obtains a feature map after passing through the feature extraction network module, and then the feature map is used as the input of the detection frame prediction module and the mask prediction module to carry out detection frame prediction and mask prediction respectively. In the original CenterNet, the feature map only outputs the coordinate type and the confidence coefficient of a detection frame of the water accumulation region after passing through a detection frame prediction module, and does not output the specific position of the water accumulation region; the mask prediction module of the invention can convert the characteristic image into a mask image and output a corresponding water accumulation area as a position point coordinate, and the mask image is consistent with the original image in size, and the content is the confidence coefficient of the water accumulation area in the original image; the higher the confidence, the higher the possibility that the water-accumulating area exists in the area of the original image. The coupling output module is used for carrying out logic judgment on output results of the detection frame prediction module and the mask prediction module, and outputting the corresponding position of the detection frame and the corresponding position coordinate in the mask image if the output confidence coefficient of the detection frame prediction module and the output confidence coefficient of the mask prediction module are both greater than a set threshold value.
And (3) training the water logging detection model by using a pyrorch training frame, adopting an SGD learning method, wherein the basic learning rate is 0.001, the weight attenuation is 0.0005, the momentum is 0.9, training 100 rounds, and reducing the learning rate by ten times in a cycle of 40 rounds.
And applying the ponding detection model to videos of different scenes, picking out a picture of an undetected ponding area and a picture of an erroneously detected ponding area, and adding the two images as positive and negative samples into a training set for retraining.
And converting the obtained single-precision detection model into a half-precision detection model through TensorRT, optimizing the scale of the model, reducing the occupancy rate of the memory and improving the detection efficiency of the accumulated water.
The camera transmits the video back to the terminal processor through the network cable, and the terminal processor analyzes the video by using the ponding detection model. Meanwhile, a ponding detection threshold value needs to be preset, and the detection frame score, the mask score and the threshold value output by the model are judged. And if the detection score and the mask score exceed the threshold value, judging that the detection frame and the mask are an effective detection frame and an effective mask, and simultaneously outputting the coordinates of the detection frame and the coordinates of the mask (namely the specific position coordinates of the water accumulation area).
The detection picture is automatically stored, the coordinates of the detection frame, the detection time and the detection address are displayed on the display screen in a text and picture mode, and therefore the detection picture is convenient for workers to browse and check.
The technical scheme of the invention can be used in any scene with cameras, including but not limited to highways, street roads, commercial districts, parks, scenic spots and the like. The video is returned only through the network cable, and the road water accumulation area can be obtained by utilizing the terminal processor to analyze the video and judge the threshold value, so that the method has stronger expansibility and portability. The false detection rate can be effectively reduced through repeated training of the wrong samples, so that the method has high reliability and stability. Meanwhile, the information is automatically stored comprehensively, and the staff can conveniently browse and check the information.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (11)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911300910.0A CN111160155A (en) | 2019-12-17 | 2019-12-17 | Method and device for detecting stagnant water |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911300910.0A CN111160155A (en) | 2019-12-17 | 2019-12-17 | Method and device for detecting stagnant water |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111160155A true CN111160155A (en) | 2020-05-15 |
Family
ID=70557542
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911300910.0A Pending CN111160155A (en) | 2019-12-17 | 2019-12-17 | Method and device for detecting stagnant water |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111160155A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113378668A (en) * | 2021-05-28 | 2021-09-10 | 上海商汤智能科技有限公司 | Method, device and equipment for determining accumulated water category and storage medium |
| CN113837164A (en) * | 2021-11-29 | 2021-12-24 | 中化学交通建设集团有限公司 | Construction site water spot detection method and device based on image recognition technology |
| CN114219788A (en) * | 2021-12-17 | 2022-03-22 | 江西洪都航空工业集团有限责任公司 | Accumulated water detection method based on video |
| CN114299457A (en) * | 2021-12-29 | 2022-04-08 | 京东方科技集团股份有限公司 | Accumulated water depth detection method and device |
| CN114581711A (en) * | 2022-03-04 | 2022-06-03 | 北京百度网讯科技有限公司 | Target object detection method, apparatus, device, storage medium, and program product |
| CN116071656A (en) * | 2023-03-06 | 2023-05-05 | 河北工业大学 | Intelligent alarm method and system for underground substation infrared image water detection and alarm |
| CN116229362A (en) * | 2023-02-21 | 2023-06-06 | 中国地质大学（武汉） | An Intelligent Waterlogging Detection and Depth Estimation Method Based on Real-time Surveillance Video |
| CN116311103A (en) * | 2023-05-10 | 2023-06-23 | 江西云眼视界科技股份有限公司 | Road area water detection method, device, medium and equipment based on incremental learning |
| CN116740577A (en) * | 2023-08-10 | 2023-09-12 | 瞳见科技有限公司 | Road ponding identification method, system, terminal and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013118606A (en) * | 2011-10-31 | 2013-06-13 | Jvc Kenwood Corp | Image processing device, image processing method, and image processing program |
| US20180211117A1 (en) * | 2016-12-20 | 2018-07-26 | Jayant Ratti | On-demand artificial intelligence and roadway stewardship system |
| CN108985322A (en) * | 2018-06-01 | 2018-12-11 | 广东电网有限责任公司 | A kind of cable tunnel ponding positioning identifying method based on ZF-Faster RCNNs |
| CN109543691A (en) * | 2018-12-27 | 2019-03-29 | 斑马网络技术有限公司 | Ponding recognition methods, device and storage medium |
| CN109670404A (en) * | 2018-11-23 | 2019-04-23 | 江苏理工学院 | A kind of road ponding image detection method for early warning based on mixed model |
| CN110135261A (en) * | 2019-04-15 | 2019-08-16 | 北京易华录信息技术股份有限公司 | A kind of method and system of trained road anomalous identification model, road anomalous identification |
| CN110569755A (en) * | 2019-08-26 | 2019-12-13 | 浙江工商大学 | A Video-Based Intelligent Waterlogging Detection Method |
-
2019
- 2019-12-17 CN CN201911300910.0A patent/CN111160155A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013118606A (en) * | 2011-10-31 | 2013-06-13 | Jvc Kenwood Corp | Image processing device, image processing method, and image processing program |
| US20180211117A1 (en) * | 2016-12-20 | 2018-07-26 | Jayant Ratti | On-demand artificial intelligence and roadway stewardship system |
| CN108985322A (en) * | 2018-06-01 | 2018-12-11 | 广东电网有限责任公司 | A kind of cable tunnel ponding positioning identifying method based on ZF-Faster RCNNs |
| CN109670404A (en) * | 2018-11-23 | 2019-04-23 | 江苏理工学院 | A kind of road ponding image detection method for early warning based on mixed model |
| CN109543691A (en) * | 2018-12-27 | 2019-03-29 | 斑马网络技术有限公司 | Ponding recognition methods, device and storage medium |
| CN110135261A (en) * | 2019-04-15 | 2019-08-16 | 北京易华录信息技术股份有限公司 | A kind of method and system of trained road anomalous identification model, road anomalous identification |
| CN110569755A (en) * | 2019-08-26 | 2019-12-13 | 浙江工商大学 | A Video-Based Intelligent Waterlogging Detection Method |
Non-Patent Citations (4)
| Title |
|---|
| KAIMING HE等: "MASK R-CNN", 《ARXIV》, pages 1 - 12 * |
| KAIWEN DUAN等: "CenterNet: Keypoint Triplets for Object Detection", 《 PROCEEDINGS OF THE IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)》, pages 6569 - 6578 * |
| 侯精明: "深度学习技术在城市内涝监测中的应用", 《中国防汛抗旱》, vol. 29, no. 8, pages 8 * |
| 段仲静等: "深度学习目标检测方法及其主流框架综述", 《激光与光电子学进展》, vol. 57, no. 12, pages 59 - 74 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113378668A (en) * | 2021-05-28 | 2021-09-10 | 上海商汤智能科技有限公司 | Method, device and equipment for determining accumulated water category and storage medium |
| CN113837164A (en) * | 2021-11-29 | 2021-12-24 | 中化学交通建设集团有限公司 | Construction site water spot detection method and device based on image recognition technology |
| CN114219788A (en) * | 2021-12-17 | 2022-03-22 | 江西洪都航空工业集团有限责任公司 | Accumulated water detection method based on video |
| CN114299457A (en) * | 2021-12-29 | 2022-04-08 | 京东方科技集团股份有限公司 | Accumulated water depth detection method and device |
| CN114581711A (en) * | 2022-03-04 | 2022-06-03 | 北京百度网讯科技有限公司 | Target object detection method, apparatus, device, storage medium, and program product |
| CN116229362A (en) * | 2023-02-21 | 2023-06-06 | 中国地质大学（武汉） | An Intelligent Waterlogging Detection and Depth Estimation Method Based on Real-time Surveillance Video |
| CN116071656A (en) * | 2023-03-06 | 2023-05-05 | 河北工业大学 | Intelligent alarm method and system for underground substation infrared image water detection and alarm |
| CN116311103A (en) * | 2023-05-10 | 2023-06-23 | 江西云眼视界科技股份有限公司 | Road area water detection method, device, medium and equipment based on incremental learning |
| CN116740577A (en) * | 2023-08-10 | 2023-09-12 | 瞳见科技有限公司 | Road ponding identification method, system, terminal and storage medium |
| CN116740577B (en) * | 2023-08-10 | 2024-01-12 | 瞳见科技有限公司 | Road ponding identification method, system, terminal and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111160155A (en) | Method and device for detecting stagnant water | |
| CN103927878B (en) | A kind of automatic shooting device for parking offense and automatically grasp shoot method | |
| CN110738121A (en) | A kind of front vehicle detection method and detection system | |
| JP2023516500A (en) | Systems and methods for image-based location determination | |
| CN110659622A (en) | Detection method, device and system for garbage dumping | |
| CN104392468A (en) | Improved visual background extraction based movement target detection method | |
| WO2018153304A1 (en) | Map road mark and road quality collection apparatus and method based on adas system | |
| CN108550258B (en) | Vehicle queuing length detection method and device, storage medium and electronic equipment | |
| CN107194396A (en) | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system | |
| CN111832410B (en) | Forward train detection method based on fusion of vision and laser radar | |
| CN108052887A (en) | A kind of doubtful illegal land automatic recognition system and method for merging SLAM/GNSS information | |
| CN111383204A (en) | Video image fusion method, fusion device, panoramic monitoring system and storage medium | |
| CN110427911B (en) | Road detection method, device, equipment and storage medium | |
| CN107146239A (en) | Satellite video moving target detection method and system | |
| CN111856445A (en) | A target detection method, device, equipment and system | |
| CN108549862A (en) | Abnormal scene detection method and device | |
| CN110794828A (en) | A landmark localization method fused with semantic information | |
| CN111563428A (en) | Airport parking space intrusion detection method and system | |
| US20130135446A1 (en) | Street view creating system and method thereof | |
| CN116030366A (en) | Power line inspection detection method and system | |
| CN103605960B (en) | A kind of method for identifying traffic status merged based on different focal video image | |
| CN111753579A (en) | Detection method and device for designated travel tool | |
| WO2024055966A1 (en) | Multi-camera target detection method and apparatus | |
| KR20190063350A (en) | Method of detecting a shooting direction and apparatuses performing the same | |
| CN107169969A (en) | A kind of measurement of highway crag collapse deposit thing size and warning system based on FPGA |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200515 |