CN117876966A - Intelligent traffic security monitoring system and method based on AI analysis - Google Patents
Intelligent traffic security monitoring system and method based on AI analysis Download PDFInfo
- Publication number
- CN117876966A CN117876966A CN202410083728.9A CN202410083728A CN117876966A CN 117876966 A CN117876966 A CN 117876966A CN 202410083728 A CN202410083728 A CN 202410083728A CN 117876966 A CN117876966 A CN 117876966A
- Authority
- CN
- China
- Prior art keywords
- target object
- semantic
- sequence
- behavior
- training
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 178
- 238000004458 analytical method Methods 0.000 title claims abstract description 77
- 238000000034 method Methods 0.000 title claims abstract description 16
- 238000001514 detection method Methods 0.000 claims abstract description 29
- 238000012545 processing Methods 0.000 claims abstract description 24
- 230000006399 behavior Effects 0.000 claims description 237
- 238000012549 training Methods 0.000 claims description 103
- 206010000117 Abnormal behaviour Diseases 0.000 claims description 31
- 238000000605 extraction Methods 0.000 claims description 26
- 238000010586 diagram Methods 0.000 claims description 23
- 239000013598 vector Substances 0.000 claims description 22
- 238000005457 optimization Methods 0.000 claims description 18
- 238000012800 visualization Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000003062 neural network model Methods 0.000 claims description 12
- 238000013527 convolutional neural network Methods 0.000 claims description 10
- 230000009466 transformation Effects 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 5
- 230000009471 action Effects 0.000 abstract description 3
- 238000007726 management method Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 10
- 230000004044 response Effects 0.000 description 8
- 238000013135 deep learning Methods 0.000 description 7
- 206010039203 Road traffic accident Diseases 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000000750 progressive effect Effects 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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; CALCULATING OR 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Image Analysis (AREA)
Abstract
The application discloses a wisdom traffic security monitoring system and method based on AI analysis, it is through wisdom traffic camera real-time supervision gather the road monitoring video, and introduce video processing and analysis algorithm based on artificial intelligence and degree of depth study at the rear end and carry out this road monitoring video's semantic analysis, thereby discernment and track vehicle, pedestrian and other object targets, and carry out the unusual action (such as retrograde, red light etc.) detection of target object, carry out the warning in time to unusual action simultaneously, so that relevant department can take measures in time. Therefore, the monitoring accuracy, the real-time performance and the efficiency of the intelligent traffic security monitoring system are improved, and the traffic security and management are better guaranteed.
Description
Technical Field
The application relates to the technical field of intelligent security monitoring, in particular to an intelligent traffic security monitoring system and method based on AI analysis.
Background
Traffic security monitoring refers to real-time monitoring and management of traffic scenes by installing cameras or other monitoring devices so as to improve traffic security and management efficiency. The traffic security monitoring system can be used for traffic places such as urban roads, highways, parking lots and the like, and helps to monitor traffic flow, identify illegal behaviors, provide evidence and the like.
However, the conventional traffic security monitoring system mainly depends on manual operation, and monitoring personnel need to observe a monitoring picture for a long time, so that fatigue and errors are easy to occur. Meanwhile, due to the limitation of the range and efficiency of manual monitoring, the comprehensive monitoring of a large-scale traffic scene cannot be realized. In addition, the traditional traffic security monitoring system identifies the illegal behaviors by means of manual observation and judgment, has limited accuracy and limited human vision and reaction capability, and is easy to miss detection or misjudgment, so that the illegal behaviors are not found and processed in time. In addition, the traditional traffic security monitoring system generally needs manual intervention to discover and process the illegal behaviors, and cannot realize real-time monitoring and timely response, so that delay processing of the illegal behaviors is caused, traffic accidents cannot be effectively prevented, and traffic management efficiency cannot be improved.
Therefore, an intelligent traffic security monitoring system based on AI analysis is desired.
Disclosure of Invention
The application provides an intelligent traffic security monitoring system and method based on AI analysis, which is characterized in that an intelligent traffic camera is used for monitoring and collecting road monitoring videos in real time, and video processing and analysis algorithms based on artificial intelligence and deep learning are introduced into the rear end to conduct semantic analysis of the road monitoring videos, so that targets of traffic tools, pedestrians and other objects are identified and tracked, abnormal behaviors (such as retrograde, red light running and the like) of the target objects are detected, and meanwhile, the abnormal behaviors are timely alarmed, so that related departments can timely take measures. Therefore, the monitoring accuracy, the real-time performance and the efficiency of the intelligent traffic security monitoring system are improved, and the traffic security and management are better guaranteed.
The application also provides an AI analysis-based intelligent traffic security monitoring system, which comprises:
the road monitoring video acquisition module is used for acquiring a road monitoring video acquired by the intelligent traffic camera;
the target object interested region extraction module is used for enabling the road monitoring video to pass through a target object detection network to obtain a target object interested region monitoring video to be detected;
the video key frame extraction module is used for extracting a sequence of the key frame of the region of interest of the target object to be detected from the monitoring video of the region of interest of the target object to be detected;
the target object behavior semantic feature extraction module is used for respectively carrying out feature extraction on the sequence of the key frames of the region of interest of the target object to be detected through a target object behavior semantic feature extractor based on a deep neural network model so as to obtain a sequence of a target object behavior semantic feature map;
the target object behavior semantic feature visualization module is used for carrying out space semantic saliency analysis on the sequence of the target object behavior semantic feature map to obtain a sequence of a target object behavior semantic saliency expression feature map;
the target object behavior semantic association coding module is used for performing behavior pattern time sequence dynamic semantic association coding on the sequence of the target object behavior semantic saliency expression feature map to obtain target object behavior dynamic semantic coding features;
And the target object abnormal behavior detection module is used for determining whether the target object to be detected has abnormal behaviors or not based on the dynamic semantic coding characteristics of the target object behaviors and determining whether an alarm signal is sent or not.
In the intelligent traffic security monitoring system based on AI analysis, the video key frame extraction module is configured to: and sparse sampling is carried out on the monitoring video of the region of interest of the target object to be detected so as to obtain a sequence of key frames of the region of interest of the target object to be detected.
In the intelligent traffic security monitoring system based on AI analysis, the deep neural network model is a convolutional neural network model.
In the intelligent traffic security monitoring system based on AI analysis, the target object behavior semantic feature display module is configured to: and passing the sequence of the target object behavior semantic feature map through a target object behavior semantic visualization based on a spatial attention layer to obtain the sequence of the target object behavior semantic saliency expression feature map.
In the intelligent traffic security monitoring system based on AI analysis, the target object behavior semantic feature display module is configured to:
processing the sequence of the target object behavior semantic feature map by using the following saliency formula in a target object behavior semantic visualization based on a spatial attention layer;
Wherein, the saliency formula is:wherein,sequence representing a semantic feature map of the behavior of said target object,/->Indicating the use of a 3 x 3 convolution kernel for the convolution processing,/->Indicating the use of a 1 x 1 convolution kernel for the convolution processing,/->Representation->Function (F)>And representing the sequence of the target object behavior semantic saliency expression feature images.
In the intelligent traffic security monitoring system based on AI analysis, the target object behavior semantic association coding module is configured to: and enabling the sequence of the target object behavior semantic significance expression feature map to pass through a behavior pattern time sequence dynamic encoder based on a converter module to obtain a target object behavior dynamic semantic encoding feature vector serving as the target object behavior dynamic semantic encoding feature.
In the intelligent traffic security monitoring system based on AI analysis, the target object abnormal behavior detection module includes: the target object behavior anomaly detection unit is used for enabling the dynamic semantic coding feature vector of the target object behavior to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the target object to be detected has an anomaly behavior or not; and the alarm signal sending unit is used for responding to the abnormal behavior of the target object to be detected and sending an alarm signal.
The intelligent traffic security monitoring system based on AI analysis further comprises a training module for training the convolutional neural network model-based target object behavior semantic feature extractor, the spatial attention layer-based target object behavior semantic visualizer, the converter module-based behavior pattern time sequence dynamic encoder and the classifier; wherein, training module includes: the training road monitoring video acquisition unit is used for acquiring training road monitoring videos acquired by the intelligent traffic cameras; the training target object interested region extraction unit is used for enabling the training road monitoring video to pass through the target object detection network to obtain a training target object interested region monitoring video to be detected; the training video key frame extraction unit is used for extracting a sequence of key frames of the region of interest of the training target object to be detected from the monitoring video of the region of interest of the training target object to be detected; the training target object behavior semantic feature extraction unit is used for respectively carrying out feature extraction on the sequence of the key frames of the region of interest of the training target object to be detected through the target object behavior semantic feature extractor based on the deep neural network model so as to obtain a sequence of a training target object behavior semantic feature map; the training target object behavior semantic feature display unit is used for carrying out space semantic saliency analysis on the sequence of the training target object behavior semantic feature map to obtain a sequence of a training target object behavior semantic saliency expression feature map; the training optimization unit is used for carrying out feature optimization on the sequence of the training target object behavior semantic saliency expression feature map so as to obtain the sequence of the optimized training target object behavior semantic saliency expression feature map; the training target object behavior semantic association coding unit is used for enabling the sequence of the optimized training target object behavior semantic saliency expression feature images to pass through the behavior pattern time sequence dynamic coder based on the converter module so as to obtain optimized training target object behavior dynamic semantic coding feature vectors; the training classification unit is used for enabling the optimized training target object behavior dynamic semantic coding feature vector to pass through the classifier to obtain a classification loss function value; the training unit is used for training the target object behavior semantic feature extractor based on the convolutional neural network model, the target object behavior semantic display based on the spatial attention layer, the behavior pattern time sequence dynamic encoder based on the converter module and the classifier based on the classification loss function value.
In the intelligent traffic security monitoring system based on AI analysis, the training optimizing unit includes: the cascading subunit is used for cascading the sequence of the training target object behavior semantic significance expression feature map into a cascading feature map along a channel; the linear transformation subunit is used for carrying out linear transformation on the cascade characteristic diagram so that the width and the height of each characteristic matrix along the channel dimension in the cascade characteristic diagram are equal to obtain a converted cascade characteristic diagram; the channel dimension optimization subunit is used for carrying out channel dimension optimization on the converted cascade feature images so as to obtain optimized cascade feature images; and the feature map restoration subunit is used for restoring the optimized cascade feature map into a sequence of feature maps to obtain a sequence of the optimized training target object behavior semantic significance expression feature map.
The application also provides an AI analysis-based intelligent traffic security monitoring method, which comprises the following steps:
acquiring a road monitoring video acquired by an intelligent traffic camera;
the road monitoring video passes through a target object detection network to obtain a target object region of interest monitoring video to be detected;
extracting a sequence of key frames of the region of interest of the target object to be detected from the region of interest monitoring video of the target object to be detected;
Respectively extracting features of the sequence of key frames of the region of interest of the target object to be detected through a target object behavior semantic feature extractor based on a deep neural network model to obtain a sequence of a target object behavior semantic feature map;
carrying out space semantic significance analysis on the sequence of the target object behavior semantic feature map to obtain a sequence of a target object behavior semantic significance expression feature map;
performing behavior pattern time sequence dynamic semantic association coding on the sequence of the target object behavior semantic saliency expression feature map to obtain target object behavior dynamic semantic coding features;
and determining whether the target object to be detected has abnormal behaviors or not based on the dynamic semantic coding features of the target object behaviors, and determining whether an alarm signal is sent or not.
Compared with the prior art, the intelligent traffic security monitoring system and the intelligent traffic security monitoring method based on AI analysis, which are provided by the application, collect the road monitoring video through the intelligent traffic camera in real time, introduce the video processing and analysis algorithm based on artificial intelligence and deep learning at the rear end to carry out the semantic analysis of the road monitoring video, so as to identify and track the targets of traffic tools, pedestrians and other objects, detect the abnormal behaviors (such as retrograde, red light running and the like) of the target objects, and alarm the abnormal behaviors in time, so that related departments can take measures in time. Therefore, the monitoring accuracy, the real-time performance and the efficiency of the intelligent traffic security monitoring system are improved, and the traffic security and management are better guaranteed.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
fig. 1 is a block diagram of an intelligent traffic security monitoring system based on AI analysis provided in an embodiment of the present application.
Fig. 2 is a flowchart of an intelligent traffic security monitoring method based on AI analysis provided in an embodiment of the present application.
Fig. 3 is a schematic diagram of a system architecture of an intelligent traffic security monitoring method based on AI analysis according to an embodiment of the present application.
Fig. 4 is an application scenario diagram of an intelligent traffic security monitoring system based on AI analysis provided in an embodiment of the present application.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the embodiments of the present application will be described in further detail with reference to the accompanying drawings. The illustrative embodiments of the present application and their description are presented herein to illustrate the application and not to limit the application.
Unless defined otherwise, all technical and scientific terms used in the examples of this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present application.
In the description of the embodiments of the present application, unless otherwise indicated and defined, the term "connected" should be construed broadly, and for example, may be an electrical connection, may be a communication between two elements, may be a direct connection, or may be an indirect connection via an intermediary, and it will be understood by those skilled in the art that the specific meaning of the term may be understood according to the specific circumstances.
It should be noted that, the term "first\second\third" in the embodiments of the present application is merely to distinguish similar objects, and does not represent a specific order for the objects, it is to be understood that "first\second\third" may interchange a specific order or sequence where allowed. It is to be understood that the "first\second\third" distinguishing objects may be interchanged where appropriate such that the embodiments of the present application described herein may be implemented in sequences other than those illustrated or described herein.
The traditional traffic security monitoring system has a plurality of defects, the traditional system mainly depends on manual operation, and monitoring personnel need to observe a monitoring picture for a long time, so that fatigue and errors are easy to occur; the range and the efficiency of manual monitoring are limited, and the comprehensive monitoring of a large-scale traffic scene cannot be realized; the traditional system recognizes the illegal behaviors by means of manual observation and judgment, has limited accuracy and is easy to miss detection or misjudge; the traditional system generally needs manual intervention to discover and process the illegal behaviors, cannot realize real-time monitoring and timely response, causes delay processing of the illegal behaviors, and cannot effectively prevent traffic accidents and improve traffic management efficiency; because of the need of manual intervention, the traditional system cannot realize real-time monitoring and timely response, and delay processing of illegal behaviors is caused. These drawbacks have resulted in the traditional traffic security monitoring systems having significant limitations in terms of monitoring efficiency, accuracy and real-time response, failing to meet the increasing traffic security management demands.
The application provides an intelligent traffic security monitored control system based on AI analysis, can realize following function: road traffic information including the number, speed, direction, license plate number and the like of vehicles is collected in real time through a high-definition camera and a sensor; through deep learning and image recognition technology, intelligent analysis is carried out on traffic information, abnormal conditions such as illegal behaviors, traffic jams, accident occurrence and the like are recognized, and timely alarming and solution providing are carried out; through cloud computing and big data technology, traffic information is counted and mined, and visual reports such as traffic flow diagrams, traffic thermodynamic diagrams, traffic prediction diagrams and the like are generated, so that scientific basis is provided for traffic management and planning; through artificial intelligence and natural language processing technology, a voice interaction function of the intelligent traffic security monitoring system is constructed, so that a user can acquire required traffic information and service through voice commands or problems.
In one embodiment of the present application, fig. 1 is a block diagram of an intelligent traffic security monitoring system based on AI analysis provided in an embodiment of the present application. As shown in fig. 1, an intelligent traffic security monitoring system 100 based on AI analysis according to an embodiment of the present application includes: the road monitoring video acquisition module 110 is used for acquiring a road monitoring video acquired by the intelligent traffic camera; the target object interested region extraction module 120 is configured to pass the road monitoring video through a target object detection network to obtain a target object interested region monitoring video to be detected; the video key frame extracting module 130 is configured to extract a sequence of key frames of the region of interest of the target object to be detected from the monitored video of the region of interest of the target object to be detected; the target object behavior semantic feature extraction module 140 is configured to perform feature extraction on the sequence of key frames of the region of interest of the target object to be detected through a target object behavior semantic feature extractor based on a deep neural network model, so as to obtain a sequence of a target object behavior semantic feature map; the target object behavior semantic feature visualization module 150 is configured to perform spatial semantic saliency analysis on the sequence of the target object behavior semantic feature map to obtain a sequence of a target object behavior semantic saliency expression feature map; the target object behavior semantic association coding module 160 is configured to perform behavior pattern time sequence dynamic semantic association coding on the sequence of the target object behavior semantic saliency expression feature map to obtain a target object behavior dynamic semantic coding feature; the target object abnormal behavior detection module 170 is configured to determine whether an abnormal behavior exists in a target object to be detected based on the dynamic semantic coding feature of the target object behavior, and determine whether to send out an alarm signal.
Aiming at the technical problems, in the technical scheme of the application, an intelligent traffic security monitoring system based on AI analysis is provided. The intelligent traffic safety monitoring system is one system utilizing artificial intelligent technology to raise traffic management efficiency and safety. The system can realize real-time monitoring, analysis and response to traffic scenes, thereby effectively preventing and reducing traffic accidents. In addition, video stream processing based on AI analysis, namely, through carrying out deep learning on video streams collected by a camera or monitoring equipment, the behavior characteristics of a target object are extracted, and whether abnormal or illegal behaviors exist is judged.
The intelligent traffic security monitoring system can realize the following functions:
video monitoring and analysis: the intelligent traffic security monitoring system acquires real-time video streams of traffic scenes through cameras or monitoring equipment, and analyzes the real-time video streams by utilizing a computer vision technology. For example, target detection and tracking algorithms are used to identify and track vehicles, pedestrians, and other objects. This may help monitor traffic flow, detect abnormal behavior (e.g., reverse, red light running, etc.), and provide real-time traffic condition information.
Traffic event detection: the intelligent traffic security monitoring system may use a deep learning algorithm to detect traffic events such as accidents, congestion, traffic violations, and the like. By analyzing video streams or other sensor data, the intelligent traffic security monitoring system can automatically identify and alarm so that relevant departments can take measures in time.
Behavioral analysis and prediction: the intelligent traffic security monitoring system can predict potential traffic hazards by analyzing the behavior patterns of traffic participants, such as the speed, acceleration, lane change behavior, etc. of vehicles. For example, the intelligent traffic security monitoring system can identify dangerous behaviors such as high-speed running, frequent lane changing or sudden braking and the like and give an alarm in time.
Real-time monitoring and alarming: the intelligent traffic security monitoring system can realize real-time monitoring and alarming functions through networking with a traffic management center or related departments. When the intelligent traffic security monitoring system detects traffic events or abnormal behaviors, the intelligent traffic security monitoring system can automatically trigger an alarm and send relevant information to related personnel so as to take measures in time.
Data analysis and decision support: the intelligent traffic security monitoring system can collect and analyze a large amount of traffic data, such as traffic flow, accident statistics, illegal behaviors and the like, and the data can be used for decision support in traffic planning, optimizing traffic signals, improving road design and the like.
Specifically, the technical concept of the application is to monitor and collect the road monitoring video in real time through the intelligent traffic camera, introduce the video processing and analysis algorithm based on artificial intelligence and deep learning at the rear end to carry out the semantic analysis of the road monitoring video, so as to identify and track the targets of vehicles, pedestrians and other objects, detect the abnormal behaviors (such as retrograde, red light running and the like) of the target objects, and simultaneously alarm the abnormal behaviors in time, so that the related departments can take measures in time. Therefore, the monitoring accuracy, the real-time performance and the efficiency of the intelligent traffic security monitoring system are improved, and the traffic security and management are better guaranteed.
Specifically, in the technical scheme of the application, firstly, a road monitoring video collected by an intelligent traffic camera is obtained. It should be appreciated that the road monitoring video typically contains a significant amount of information, including background interference information. Therefore, in order to improve the monitoring efficiency and accuracy of the system, in the technical scheme of the application, the road monitoring video needs to be passed through the target object detection network to obtain the monitoring video of the region of interest of the target object to be detected. In particular, in a traffic scenario, the target object may be various vehicles (e.g., automobiles, motorcycles, bicycles, etc.), pedestrians, and the like. Through the processing of the target object detection network, the system can automatically identify and locate the target object in the video and extract the region of interest where the target object is located. In addition, the application of the target object detection network can also help to filter irrelevant background information in the monitoring video, and only the interested area related to the target object is reserved, so that the calculation amount of subsequent processing can be reduced, and the recognition accuracy of the illegal behaviors is improved.
It should be understood that, in the traffic security monitoring system, the monitoring video of the region of interest of the target object to be detected is a continuous video stream, and includes a large number of frame images. However, processing each frame of image of a continuous video stream may result in excessive computation, reducing the real-time and efficiency of the system. Therefore, in order to reduce the data size and extract the behavior semantic key feature information of the target object, in the technical scheme of the application, sparse sampling is required to be performed on the monitored video of the region of interest of the target object to be detected so as to obtain the sequence of key frames of the region of interest of the target object to be detected. By means of sparse sampling, a part of key frame images can be selected from a continuous video stream to be processed and analyzed, and the key frame images generally contain important action and behavior information of a target object and can represent behavior semantic features of the whole video about the target object.
In a specific embodiment of the present application, the video key frame extraction module is configured to: and sparse sampling is carried out on the monitoring video of the region of interest of the target object to be detected so as to obtain a sequence of key frames of the region of interest of the target object to be detected.
And then, the sequence of the target object interesting region key frames to be detected is subjected to feature mining in a target object behavior semantic feature extractor based on a convolutional neural network model, so that behavior semantic feature information about the target object in each target object interesting region key frame to be detected is extracted, and a sequence of a target object behavior semantic feature map is obtained.
Specifically, the deep neural network model is a convolutional neural network model.
Furthermore, in the intelligent traffic security monitoring system, the behavior semantic feature information about the target object is contained in the behavior semantic feature map of the target object, but some redundant or unimportant feature information exists at the same time. Therefore, in order to highlight the behavior characteristics of the target object and improve the accuracy and the interpretability of behavior analysis, in the technical scheme of the application, the sequence of the behavior semantic feature map of the target object is further processed through a target object behavior semantic display based on a spatial attention layer to obtain the sequence of the behavior semantic saliency expression feature map of the target object. The target object behavior semantic feature map can be weighted by using the target object behavior semantic display based on the spatial attention layer to highlight key parts related to the target object behavior, and redundant or unimportant information is restrained, so that the system focuses on important details of the target object behavior, and the system is helpful for better understanding and judging whether the target object has abnormal behavior.
In a specific embodiment of the present application, the target object behavior semantic feature visualization module is configured to: and passing the sequence of the target object behavior semantic feature map through a target object behavior semantic visualization based on a spatial attention layer to obtain the sequence of the target object behavior semantic saliency expression feature map.
Further, the target object behavior semantic feature display module is used for: processing the sequence of the target object behavior semantic feature map by using the following saliency formula in a target object behavior semantic visualization based on a spatial attention layer; wherein, the saliency formula is:wherein (1)>Sequence representing a semantic feature map of the behavior of said target object,/->Indicating the use of a 3 x 3 convolution kernel for the convolution processing,/->Indicating the use of a 1 x 1 convolution kernel for the convolution processing,/->Representation->Function (F)>And representing the sequence of the target object behavior semantic saliency expression feature images.
Then, the sequence of the target object behavior semantic significance expression feature map reflects the key behavior semantic features of the target object in a period of time in the intelligent traffic security monitoring system. And, it is also considered that these key behavior semantic features about the target object have an association relationship based on a time dimension. Therefore, in order to more comprehensively and fully understand and describe the behavior semantic time sequence dynamic characteristics of the target object so as to better represent the behavior mode of the target object, in the technical scheme of the application, the sequence of the behavior semantic significance expression characteristic diagram of the target object is further processed by a behavior mode time sequence dynamic encoder based on a converter module so as to obtain the behavior dynamic semantic coding characteristic vector of the target object. Particularly, the dynamic correlation characteristic information of the behavior semantics of the target object in the time dimension can be effectively extracted by using the behavior pattern time sequence dynamic encoder based on the converter module to encode, so that the dynamic pattern of the behavior of the target object can be reflected, different behavior patterns such as acceleration, deceleration, lane change and the like can be better analyzed and identified, and a more accurate basis is provided for subsequent behavior anomaly detection.
In a specific embodiment of the present application, the target object behavior semantic association encoding module is configured to: and enabling the sequence of the target object behavior semantic significance expression feature map to pass through a behavior pattern time sequence dynamic encoder based on a converter module to obtain a target object behavior dynamic semantic encoding feature vector serving as the target object behavior dynamic semantic encoding feature.
And then, the dynamic semantic coding feature vector of the target object behavior passes through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the target object to be detected has abnormal behavior or not. That is, the time sequence dynamic association characteristic information of the behavior semantics of the target object in the time dimension is utilized to perform classification processing, so that abnormal behaviors (such as retrograde, red light running and the like) of the target object are detected, and an alarm signal is sent out in response to the existence of the abnormal behaviors of the target object to be detected. Therefore, the monitoring accuracy, the real-time performance and the efficiency of the intelligent traffic security monitoring system are improved, and the traffic security and management are better guaranteed.
In a specific embodiment of the present application, the target object abnormal behavior detection module includes: the target object behavior anomaly detection unit is used for enabling the dynamic semantic coding feature vector of the target object behavior to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the target object to be detected has an anomaly behavior or not; and the alarm signal sending unit is used for responding to the abnormal behavior of the target object to be detected and sending an alarm signal.
The behavior of the target object can be automatically monitored by passing the dynamic semantic coding feature vector of the behavior of the target object through the classifier, and a monitoring picture is not required to be observed manually for a long time, so that the burden of monitoring personnel is reduced, and fatigue and errors are avoided; the comprehensive monitoring of a large-scale traffic scene can be realized, and the limitations of the traditional manual monitoring range and efficiency are overcome.
The target object behavior dynamic semantic coding feature vector is subjected to abnormal behavior detection through the classifier, so that the recognition accuracy can be improved, errors caused by manual judgment are reduced, and the conditions of missed detection and misjudgment are reduced. The classifier is adopted for classification, real-time monitoring and timely response can be realized, and manual intervention is not needed, so that delay processing of illegal behaviors is avoided, traffic accidents are effectively prevented, and traffic management efficiency is improved.
In one embodiment of the present application, the intelligent traffic security monitoring system based on AI analysis further includes a training module for training the target object behavior semantic feature extractor based on convolutional neural network model, the target object behavior semantic display based on spatial attention layer, the behavior pattern time sequence dynamic encoder based on converter module, and the classifier; wherein, training module includes: the training road monitoring video acquisition unit is used for acquiring training road monitoring videos acquired by the intelligent traffic cameras; the training target object interested region extraction unit is used for enabling the training road monitoring video to pass through the target object detection network to obtain a training target object interested region monitoring video to be detected; the training video key frame extraction unit is used for extracting a sequence of key frames of the region of interest of the training target object to be detected from the monitoring video of the region of interest of the training target object to be detected; the training target object behavior semantic feature extraction unit is used for respectively carrying out feature extraction on the sequence of the key frames of the region of interest of the training target object to be detected through the target object behavior semantic feature extractor based on the deep neural network model so as to obtain a sequence of a training target object behavior semantic feature map; the training target object behavior semantic feature display unit is used for carrying out space semantic saliency analysis on the sequence of the training target object behavior semantic feature map to obtain a sequence of a training target object behavior semantic saliency expression feature map; the training optimization unit is used for carrying out feature optimization on the sequence of the training target object behavior semantic saliency expression feature map so as to obtain the sequence of the optimized training target object behavior semantic saliency expression feature map; the training target object behavior semantic association coding unit is used for enabling the sequence of the optimized training target object behavior semantic saliency expression feature images to pass through the behavior pattern time sequence dynamic coder based on the converter module so as to obtain optimized training target object behavior dynamic semantic coding feature vectors; the training classification unit is used for enabling the optimized training target object behavior dynamic semantic coding feature vector to pass through the classifier to obtain a classification loss function value; the training unit is used for training the target object behavior semantic feature extractor based on the convolutional neural network model, the target object behavior semantic display based on the spatial attention layer, the behavior pattern time sequence dynamic encoder based on the converter module and the classifier based on the classification loss function value.
In the technical scheme of the application, after sparse sampling is performed on the monitoring video of the region of interest of the training target object to be detected, the obtained sequence of key frames of the region of interest of the training target object to be detected weakens the expression of the semantic time sequence relevance of the source video image frames of the monitoring video of the region of interest of the training target object to be detected, so that after the sequence of the training target object behavior semantic feature map passes through the target object behavior semantic visualization based on the spatial attention layer, the correlation among the obtained sequences of the training target object behavior semantic saliency expression feature map is further reduced due to strengthening of the local spatial image semantic features under the spatial distribution of the image semantic features, and even if the sequence of the training target object behavior semantic saliency expression feature map is subjected to context relevance based on the image semantic feature context of the source video image frames of the training target object to be detected through the behavior pattern time sequence dynamic encoder based on the converter module, the overall feature distribution sparsity of the obtained training target object behavior dynamic semantic coding feature vector is difficult to be eliminated.
Therefore, the applicant of the application considers to promote the channel dimension feature distribution integrity of the sequence of the training target object behavior semantic significance expression feature map as a whole to inhibit the overall feature distribution sparsity of the training target object behavior dynamic semantic coding feature vector, so as to improve the convergence effect when the training target object behavior dynamic semantic coding feature vector converges class probability through a classifier.
Therefore, the applicant of the present application preferably first concatenates the sequence of the training target object behavior semantic significance expression feature map into a cascade feature map along a channel, then performs linear transformation on the cascade feature map so as to make the width and the height of the feature matrix equal, and then performs channel dimension optimization on the transformed cascade feature map. The training optimization unit comprises: the cascading subunit is used for cascading the sequence of the training target object behavior semantic significance expression feature map into a cascading feature map along a channel; the linear transformation subunit is used for carrying out linear transformation on the cascade characteristic diagram so that the width and the height of each characteristic matrix along the channel dimension in the cascade characteristic diagram are equal to obtain a converted cascade characteristic diagram;
the channel dimension optimization subunit is used for carrying out channel dimension optimization on the converted cascade feature images so as to obtain optimized cascade feature images; and the feature map restoration subunit is used for restoring the optimized cascade feature map into a sequence of feature maps to obtain a sequence of the optimized training target object behavior semantic significance expression feature map. The channel dimension optimization is carried out on the converted cascade feature diagram by the following optimization formula so as to obtain an optimized cascade feature diagram; wherein, the optimization formula is: Wherein (1)>And->The first part of the converted cascade characteristic diagram is the part +.>And->Feature matrix of position, and->Is a scale-regulated superparameter,/->Is the +.sup.th in channel direction of the converted cascade feature map>Transposed matrix of feature matrix of locations, +.>Is the +.sup.th in channel direction of the converted cascade feature map>Inverse of the feature matrix of the location, +.>Is the first of the optimized cascade feature diagram along the channel directionFeature matrix of the location>Representing addition by position +.>Representing multiplication by location +.>Representing a matrix multiplication.
Here, the progressive structured embedding calculation of the feature matrix with channel adjacent distribution of the cascade feature map is used for predicting the coupling distribution direction of the local feature distribution of the cascade feature map along the channel in the high-dimensional feature space, so that the transmission pattern representation generated based on iteration of channel coupling is determined based on the distribution progressive center, the context relation of the cascade feature map based on the scene layout of the feature matrix is reconstructed in a mode of refining the projection standardization proposal of the cascade feature map from bottom to top along the channel dimension, and the channel dimension integrity of the feature representation of the cascade feature map is improved, so that the overall feature distribution sparsity of the dynamic semantic coding feature vector of the target object behavior is restrained, and the convergence effect of the dynamic semantic coding feature vector is improved when the dynamic feature vector is converged through a classifier, namely the training efficiency and the accuracy of classification results are improved. Therefore, the intelligent traffic security monitoring system can be utilized to automatically identify and track traffic tools, pedestrians and other object targets, abnormal behaviors of the target objects are detected, and meanwhile, the abnormal behaviors are timely alarmed, so that relevant departments can timely take measures, and the monitoring accuracy, instantaneity and efficiency of the intelligent traffic security monitoring system are improved, so that traffic safety and management are better guaranteed.
In summary, the intelligent traffic security monitoring system 100 based on AI analysis according to the embodiments of the present application is illustrated, which can realize real-time monitoring, analysis and response to traffic scenes, so as to effectively prevent and reduce occurrence of traffic accidents. In addition, video stream processing based on AI analysis, namely, through carrying out deep learning on video streams collected by a camera or monitoring equipment, the behavior characteristics of a target object are extracted, and whether abnormal or illegal behaviors exist is judged.
As described above, the intelligent transportation security monitoring system 100 based on AI analysis according to the embodiment of the present application may be implemented in various terminal devices, for example, a server or the like for intelligent transportation security monitoring based on AI analysis. In one example, the intelligent traffic safety monitoring system 100 based on AI analysis according to an embodiment of the present application may be integrated into a terminal device as one software module and/or hardware module. For example, the AI-analysis-based intelligent traffic security monitoring system 100 may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the intelligent traffic safety monitoring system 100 based on AI analysis can also be one of a plurality of hardware modules of the terminal device.
Alternatively, in another example, the AI-analysis-based intelligent transportation security monitoring system 100 and the terminal device may be separate devices, and the AI-analysis-based intelligent transportation security monitoring system 100 may be connected to the terminal device through a wired and/or wireless network and transmit interactive information in a contracted data format.
Fig. 2 is a flowchart of an intelligent traffic security monitoring method based on AI analysis provided in an embodiment of the present application. Fig. 3 is a schematic diagram of a system architecture of an intelligent traffic security monitoring method based on AI analysis according to an embodiment of the present application. As shown in fig. 2 and 3, an intelligent traffic security monitoring method based on AI analysis includes: 210, acquiring a road monitoring video acquired by an intelligent traffic camera; 220, passing the road monitoring video through a target object detection network to obtain a region-of-interest monitoring video of a target object to be detected; 230, extracting a sequence of key frames of the region of interest of the target object to be detected from the monitoring video of the region of interest of the target object to be detected; 240, respectively extracting features of the sequence of key frames of the region of interest of the target object to be detected through a target object behavior semantic feature extractor based on a deep neural network model to obtain a sequence of a target object behavior semantic feature map; 250, performing spatial semantic saliency analysis on the sequence of the target object behavior semantic feature map to obtain a sequence of a target object behavior semantic saliency expression feature map; 260, performing behavior pattern time sequence dynamic semantic association coding on the sequence of the target object behavior semantic saliency expression feature map to obtain target object behavior dynamic semantic coding features; 270, determining whether the target object to be detected has abnormal behavior based on the dynamic semantic coding features of the target object behavior, and determining whether to send out an alarm signal.
It will be appreciated by those skilled in the art that the specific operations of the respective steps in the above-described intelligent traffic security monitoring method based on AI analysis have been described in detail in the above description of the intelligent traffic security monitoring system based on AI analysis with reference to fig. 1, and thus, repetitive descriptions thereof will be omitted.
Fig. 4 is an application scenario diagram of an intelligent traffic security monitoring system based on AI analysis provided in an embodiment of the present application. As shown in fig. 4, in the application scenario, first, a road monitoring video acquired by a smart traffic camera is acquired (e.g., C as illustrated in fig. 4); then, the acquired road monitoring video is input to a server (e.g., S as illustrated in fig. 4) deployed with an AI-analysis-based intelligent traffic security monitoring algorithm, wherein the server is capable of processing the road monitoring video based on the AI-analysis-based intelligent traffic security monitoring algorithm to determine whether an abnormal behavior exists in the target object to be detected and to determine whether an alarm signal is issued.
The foregoing embodiments have been provided for the purpose of illustrating the general principles of the present application and are not meant to limit the scope of the invention, but to limit the scope of the invention.
Claims (10)
1. An intelligent traffic security monitored control system based on AI analysis, characterized by comprising:
the road monitoring video acquisition module is used for acquiring a road monitoring video acquired by the intelligent traffic camera;
the target object interested region extraction module is used for enabling the road monitoring video to pass through a target object detection network to obtain a target object interested region monitoring video to be detected;
the video key frame extraction module is used for extracting a sequence of the key frame of the region of interest of the target object to be detected from the monitoring video of the region of interest of the target object to be detected;
the target object behavior semantic feature extraction module is used for respectively carrying out feature extraction on the sequence of the key frames of the region of interest of the target object to be detected through a target object behavior semantic feature extractor based on a deep neural network model so as to obtain a sequence of a target object behavior semantic feature map;
the target object behavior semantic feature visualization module is used for carrying out space semantic saliency analysis on the sequence of the target object behavior semantic feature map to obtain a sequence of a target object behavior semantic saliency expression feature map;
the target object behavior semantic association coding module is used for performing behavior pattern time sequence dynamic semantic association coding on the sequence of the target object behavior semantic saliency expression feature map to obtain target object behavior dynamic semantic coding features;
And the target object abnormal behavior detection module is used for determining whether the target object to be detected has abnormal behaviors or not based on the dynamic semantic coding characteristics of the target object behaviors and determining whether an alarm signal is sent or not.
2. The AI-analysis-based intelligent traffic security monitoring system of claim 1, wherein the video key frame extraction module is configured to: and sparse sampling is carried out on the monitoring video of the region of interest of the target object to be detected so as to obtain a sequence of key frames of the region of interest of the target object to be detected.
3. The AI analysis-based intelligent traffic security monitoring system of claim 2, wherein the deep neural network model is a convolutional neural network model.
4. The AI-analysis-based intelligent traffic security monitoring system of claim 3, wherein the target object behavior semantic feature visualization module is configured to: and passing the sequence of the target object behavior semantic feature map through a target object behavior semantic visualization based on a spatial attention layer to obtain the sequence of the target object behavior semantic saliency expression feature map.
5. The AI-analysis-based intelligent traffic security monitoring system of claim 4, wherein the target object behavior semantic feature visualization module is configured to:
Processing the sequence of the target object behavior semantic feature map by using the following saliency formula in a target object behavior semantic visualization based on a spatial attention layer;
wherein, the saliency formula is:wherein,sequence representing a semantic feature map of the behavior of said target object,/->Indicating the use of a 3 x 3 convolution kernel for the convolution processing,/->Indicating the use of a 1 x 1 convolution kernel for the convolution processing,/->Representation->Function (F)>And representing the sequence of the target object behavior semantic saliency expression feature images.
6. The AI-analysis-based intelligent traffic security monitoring system of claim 5, wherein the target object behavior semantic association encoding module is configured to: and enabling the sequence of the target object behavior semantic significance expression feature map to pass through a behavior pattern time sequence dynamic encoder based on a converter module to obtain a target object behavior dynamic semantic encoding feature vector serving as the target object behavior dynamic semantic encoding feature.
7. The AI-analysis-based intelligent traffic security monitoring system of claim 6, wherein the target object abnormal behavior detection module comprises:
The target object behavior anomaly detection unit is used for enabling the dynamic semantic coding feature vector of the target object behavior to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the target object to be detected has an anomaly behavior or not;
and the alarm signal sending unit is used for responding to the abnormal behavior of the target object to be detected and sending an alarm signal.
8. The AI-analysis-based intelligent traffic security monitoring system of claim 7, further comprising a training module for training the convolutional neural network model-based target object behavior semantic feature extractor, the spatial attention layer-based target object behavior semantic visualizer, the converter module-based behavior pattern timing dynamic encoder, and the classifier;
wherein, training module includes:
the training road monitoring video acquisition unit is used for acquiring training road monitoring videos acquired by the intelligent traffic cameras;
the training target object interested region extraction unit is used for enabling the training road monitoring video to pass through the target object detection network to obtain a training target object interested region monitoring video to be detected;
The training video key frame extraction unit is used for extracting a sequence of key frames of the region of interest of the training target object to be detected from the monitoring video of the region of interest of the training target object to be detected;
the training target object behavior semantic feature extraction unit is used for respectively carrying out feature extraction on the sequence of the key frames of the region of interest of the training target object to be detected through the target object behavior semantic feature extractor based on the deep neural network model so as to obtain a sequence of a training target object behavior semantic feature map;
the training target object behavior semantic feature display unit is used for carrying out space semantic saliency analysis on the sequence of the training target object behavior semantic feature map to obtain a sequence of a training target object behavior semantic saliency expression feature map;
the training optimization unit is used for carrying out feature optimization on the sequence of the training target object behavior semantic saliency expression feature map so as to obtain the sequence of the optimized training target object behavior semantic saliency expression feature map;
the training target object behavior semantic association coding unit is used for enabling the sequence of the optimized training target object behavior semantic saliency expression feature images to pass through the behavior pattern time sequence dynamic coder based on the converter module so as to obtain optimized training target object behavior dynamic semantic coding feature vectors;
The training classification unit is used for enabling the optimized training target object behavior dynamic semantic coding feature vector to pass through the classifier to obtain a classification loss function value;
the training unit is used for training the target object behavior semantic feature extractor based on the convolutional neural network model, the target object behavior semantic display based on the spatial attention layer, the behavior pattern time sequence dynamic encoder based on the converter module and the classifier based on the classification loss function value.
9. The AI-analysis-based intelligent traffic security monitoring system of claim 8, wherein the training optimization unit comprises:
the cascading subunit is used for cascading the sequence of the training target object behavior semantic significance expression feature map into a cascading feature map along a channel;
the linear transformation subunit is used for carrying out linear transformation on the cascade characteristic diagram so that the width and the height of each characteristic matrix along the channel dimension in the cascade characteristic diagram are equal to obtain a converted cascade characteristic diagram;
the channel dimension optimization subunit is used for carrying out channel dimension optimization on the converted cascade feature images so as to obtain optimized cascade feature images;
And the feature map restoration subunit is used for restoring the optimized cascade feature map into a sequence of feature maps to obtain a sequence of the optimized training target object behavior semantic significance expression feature map.
10. An intelligent traffic security monitoring method based on AI analysis is characterized by comprising the following steps:
acquiring a road monitoring video acquired by an intelligent traffic camera;
the road monitoring video passes through a target object detection network to obtain a target object region of interest monitoring video to be detected;
extracting a sequence of key frames of the region of interest of the target object to be detected from the region of interest monitoring video of the target object to be detected;
respectively extracting features of the sequence of key frames of the region of interest of the target object to be detected through a target object behavior semantic feature extractor based on a deep neural network model to obtain a sequence of a target object behavior semantic feature map;
carrying out space semantic significance analysis on the sequence of the target object behavior semantic feature map to obtain a sequence of a target object behavior semantic significance expression feature map;
performing behavior pattern time sequence dynamic semantic association coding on the sequence of the target object behavior semantic saliency expression feature map to obtain target object behavior dynamic semantic coding features;
And determining whether the target object to be detected has abnormal behaviors or not based on the dynamic semantic coding features of the target object behaviors, and determining whether an alarm signal is sent or not.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410083728.9A CN117876966A (en) | 2024-01-19 | 2024-01-19 | Intelligent traffic security monitoring system and method based on AI analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410083728.9A CN117876966A (en) | 2024-01-19 | 2024-01-19 | Intelligent traffic security monitoring system and method based on AI analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117876966A true CN117876966A (en) | 2024-04-12 |
Family
ID=90581083
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410083728.9A Pending CN117876966A (en) | 2024-01-19 | 2024-01-19 | Intelligent traffic security monitoring system and method based on AI analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117876966A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118135508A (en) * | 2024-05-08 | 2024-06-04 | 东揽(南京)智能科技有限公司 | Holographic traffic intersection sensing system and method based on machine vision |
| CN118135800A (en) * | 2024-05-06 | 2024-06-04 | 东揽(南京)智能科技有限公司 | Abnormal traffic event accurate identification warning method based on deep learning |
-
2024
- 2024-01-19 CN CN202410083728.9A patent/CN117876966A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118135800A (en) * | 2024-05-06 | 2024-06-04 | 东揽(南京)智能科技有限公司 | Abnormal traffic event accurate identification warning method based on deep learning |
| CN118135508A (en) * | 2024-05-08 | 2024-06-04 | 东揽(南京)智能科技有限公司 | Holographic traffic intersection sensing system and method based on machine vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102197946B1 (en) | object recognition and counting method using deep learning artificial intelligence technology | |
| US11380105B2 (en) | Identification and classification of traffic conflicts | |
| CN117876966A (en) | Intelligent traffic security monitoring system and method based on AI analysis | |
| CN111311918B (en) | Traffic management method and device based on visual analysis | |
| CN111241343A (en) | Road information monitoring and analyzing detection method and intelligent traffic control system | |
| KR20100119476A (en) | An outomatic sensing system for traffic accident and method thereof | |
| KR102122850B1 (en) | Solution for analysis road and recognition vehicle license plate employing deep-learning | |
| Xie et al. | Development of a comprehensive framework for video-based safety assessment | |
| CN113033275B (en) | Vehicle lane-changing non-turn signal lamp analysis system based on deep learning | |
| Abdel-Aty et al. | Using closed-circuit television cameras to analyze traffic safety at intersections based on vehicle key points detection | |
| CN112528759A (en) | Traffic violation behavior detection method based on computer vision | |
| CN210721850U (en) | Highway road conditions intellectual detection system | |
| CN114898297A (en) | Non-motor vehicle illegal behavior determination method based on target detection and target tracking | |
| JP2023524623A (en) | Method, apparatus, equipment, storage medium and program for detecting traffic anomalies | |
| Kim et al. | Visual extensions and anomaly detection in the pNEUMA experiment with a swarm of drones | |
| Huang et al. | Enabling versatile analysis of large scale traffic video data with deep learning and HiveQL | |
| CN116311166A (en) | Traffic obstacle recognition method and device and electronic equipment | |
| CN111627224A (en) | Vehicle speed abnormality detection method, device, equipment and storage medium | |
| Neto et al. | Computer-vision-based surveillance of intelligent transportation systems | |
| CN113850112A (en) | Road condition identification method and system based on twin neural network | |
| CN113989731A (en) | Information detection method, computing device and storage medium | |
| Song et al. | Method of Vehicle Behavior Analysis for Real-Time Video Streaming Based on Mobilenet-YOLOV4 and ERFNET | |
| Raj et al. | An Embedded Deep Learning Based Traffic Advisory System | |
| Fu | Construction site safety helmet wearing detection method based on improved YOLOv5 | |
| CN118135800B (en) | Abnormal traffic event accurate identification warning method based on deep learning |
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 |