CN112949359A - Convolutional neural network-based abnormal behavior identification method and device - Google Patents
Convolutional neural network-based abnormal behavior identification method and device Download PDFInfo
- Publication number
- CN112949359A CN112949359A CN201911262520.9A CN201911262520A CN112949359A CN 112949359 A CN112949359 A CN 112949359A CN 201911262520 A CN201911262520 A CN 201911262520A CN 112949359 A CN112949359 A CN 112949359A
- Authority
- CN
- China
- Prior art keywords
- abnormal behavior
- abnormal
- module
- detection
- video
- 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; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Abstract
The embodiment of the disclosure discloses an abnormal behavior recognition device based on a convolutional neural network, comprising: the abnormal behavior detection module is configured to acquire a video to be identified, detect a current image in the video to be identified, determine whether suspicious abnormal behaviors exist in the current image by using an abnormal behavior detection model, and when the suspicious abnormal behaviors exist in the current image, determine the category of the suspicious abnormal behaviors and a detection frame aiming at the suspicious abnormal behaviors, wherein the detection frame covers an area where the suspicious abnormal behaviors occur.
Description
Technical Field
The present disclosure relates to the field of monitoring, and in particular, to a method and an apparatus for identifying abnormal behavior based on a convolutional neural network.
Background
The stability of the public safety concerned countries and the society directly influences the life and property safety of people. In order to guarantee public safety, it is very necessary to find abnormal behaviors in a security inspection scene in real time and give early warning in time to guarantee the safety of people going out. Compared with the traditional method, the method can detect abnormal behaviors in the image and identify continuous actions by utilizing the deep learning technology. However, how to effectively, quickly, and in real time identify behavioral actions in videos remains a core task for many researchers.
In recent years, a behavior recognition algorithm based on deep learning has made a major breakthrough, and mainly the recognition rate of the convolutional neural network in the field of machine vision is greatly improved. The existing video behavior identification method in some deep learning fields can be divided into two directions, one is that a region of interest (ROI) needs to be manually selected, only a specific region in a video scene is monitored, and the whole scene cannot be effectively monitored; the other is to identify the behaviors in the whole scene, but because a 3D convolutional neural network is needed, continuous multi-frame images are detected and classified each time, so that a large amount of computer resources are consumed, the calculation time is long, and the real-time detection and classification are difficult to achieve.
Therefore, a method for detecting and identifying abnormal behaviors such as left-over parcels, barrier deliveries, fighting and the like in a video without selecting an ROI (region of interest) in the video is needed, namely, the abnormal behaviors occurring in monitoring videos of places such as public transportation, major event activities, important regions, dense people regions and the like are automatically detected, accurately positioned and identified in real time.
BRIEF SUMMARY OF THE PRESENT DISCLOSURE
According to an aspect of the embodiments of the present disclosure, there is provided an abnormal behavior recognition apparatus based on a convolutional neural network, including:
the abnormal behavior detection module is configured to acquire a video to be identified, detect a current image in the video to be identified, determine whether suspicious abnormal behaviors exist in the current image by using an abnormal behavior detection model, and when the suspicious abnormal behaviors exist in the current image, determine the category of the suspicious abnormal behaviors and a detection frame aiming at the suspicious abnormal behaviors, wherein the detection frame covers an area where the suspicious abnormal behaviors occur.
In one embodiment, the abnormal behavior detection module is further configured to intercept a predetermined number of consecutive images in the video to be identified starting from the current image using the detection box; and is
The abnormal behavior recognizing apparatus further includes: an abnormal behavior recognition module configured to acquire the predetermined number of consecutive images and the detection frame from the abnormal behavior detection module, determine whether an abnormal behavior exists in the consecutive images using an abnormal behavior recognition model, and determine a category of the abnormal behavior when it is determined that the abnormal behavior exists in the consecutive images.
In one embodiment, the abnormal behavior recognizing apparatus further includes:
a receiving module configured to receive the video to be identified and send the video to be identified to the abnormal behavior detection module; and
an output module configured to acquire and output the category of the suspicious abnormal behavior and the detection box from the abnormal behavior determination module.
In one embodiment, the output module is further configured to acquire and output the category of the abnormal behavior and the detection box from the abnormal behavior recognition module.
In one embodiment, the detection box is a rectangular box.
In one embodiment, the predetermined number is 16 and the predetermined number of consecutive images are truncated at predetermined time intervals.
In one embodiment, the abnormal behavior detection model is constructed using an abnormal behavior data set with a 2D convolutional neural network.
In one embodiment, the abnormal behavior recognition model is constructed using a 3D convolutional neural network using training data that is a succession of images that are cut from a video using the constructed abnormal behavior detection model.
In one embodiment, the 2D convolutional neural network is a YOLOV3 target detection network, where the YOLOV3 network is stacked with convolutions of 3x3 and 1x1, contains no pooling layers, and uses convolutional layer systolic pictures.
In one embodiment, the 3D convolutional neural network is based on an vgg16 framework, where 2D convolutional and pooling layers are replaced with 3D convolutional and 3D pooling layers, and the dimensions of the input layers are 16x3x112x 112.
According to another aspect of the embodiments of the present disclosure, there is provided an abnormal behavior identification method based on a convolutional neural network, including:
acquiring a video to be identified by an abnormal behavior detection module;
detecting a current image in the video to be identified by the abnormal behavior detection module;
determining, by the abnormal behavior detection module, whether there is suspicious abnormal behavior in the current image using an abnormal behavior detection model; and
when the abnormal behavior detection module determines that suspicious abnormal behaviors exist in the current image, the abnormal behavior detection module determines the class of the suspicious abnormal behaviors and a detection frame aiming at the suspicious abnormal behaviors, wherein the detection frame covers the area where the suspicious abnormal behaviors occur.
In one embodiment, the abnormal behavior recognition method further includes:
intercepting, by the abnormal behavior detection module, a predetermined number of consecutive images in the video to be recognized starting from the current image using the detection frame;
acquiring, by an abnormal behavior recognition module, the predetermined number of consecutive images and the detection frame from the abnormal behavior detection module;
determining, by the abnormal behavior recognition module, whether there is an abnormal behavior in the continuous image using an abnormal behavior recognition model; and
when the abnormal behavior identification module determines that abnormal behavior exists in the continuous images, determining the category of the abnormal behavior by the abnormal behavior identification module.
In one embodiment, the abnormal behavior recognition method further includes:
receiving the video to be identified by a receiving module and sending the video to be identified to the abnormal behavior detection module; and
obtaining and outputting, by an output module, the category of the suspicious abnormal behavior and the detection box from the abnormal behavior determination module.
In one embodiment, the abnormal behavior recognition method further includes:
and acquiring and outputting the category of the abnormal behavior and the detection frame from the abnormal behavior identification module by the output module.
In one embodiment, the detection box is a rectangular box.
In one embodiment, the predetermined number is 16 and the predetermined number of consecutive images are truncated at predetermined time intervals.
In one embodiment, the abnormal behavior detection model is constructed using an abnormal behavior data set with a 2D convolutional neural network.
In one embodiment, the abnormal behavior recognition model is constructed using a 3D convolutional neural network using training data that is a succession of images that are cut from a video using the constructed abnormal behavior detection model.
In one embodiment, the 2D convolutional neural network is a YOLOV3 target detection network, where the YOLOV3 network is stacked with convolutions of 3x3 and 1x1, contains no pooling layers, and uses convolutional layer systolic pictures.
In one embodiment, the 3D convolutional neural network is based on an vgg16 framework, where the 2D convolutional and pooling layers are changed to 3D convolutional and 3D pooling layers, and the input layer is instead 16x3x112x 112.
Drawings
The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments of the present disclosure with reference to the accompanying drawings, in which:
fig. 1 shows a schematic diagram of an abnormal behavior recognition apparatus based on a convolutional neural network according to an embodiment of the present disclosure;
FIG. 2 shows an architectural diagram of a 2D convolutional neural network, according to an embodiment of the present disclosure;
FIG. 3 shows an architectural diagram of a 3D convolutional neural network, according to an embodiment of the present disclosure;
FIG. 4 illustrates a flow diagram of a convolutional neural network-based abnormal behavior identification method, in accordance with an embodiment of the present disclosure;
FIG. 5 illustrates a flow diagram of another method of anomalous behavior identification based on convolutional neural networks, in accordance with an embodiment of the present disclosure; and
fig. 6 shows a schematic diagram of a convolutional neural network-based abnormal behavior identification system, according to an embodiment of the present disclosure.
The figures do not show all of the circuitry or structures of the embodiments. The same reference numbers will be used throughout the drawings to refer to the same or like parts or features.
Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The words "a", "an" and "the" and the like as used herein are also intended to include the meanings of "a plurality" and "the" unless the context clearly dictates otherwise. Furthermore, the terms "comprises," "comprising," or the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.
Fig. 1 shows a schematic diagram of an abnormal behavior recognition apparatus 100 based on a convolutional neural network according to an embodiment of the present disclosure. The abnormal behavior recognition apparatus 100 may include an abnormal behavior detection module 110.
The abnormal behavior detection module 110 may be configured to acquire a video to be recognized, detect a current image in the video to be recognized, determine whether suspicious abnormal behavior (e.g., a left-over package, a barrier, a fighting, etc.) exists in the current image using an abnormal behavior detection model, and when it is determined that suspicious abnormal behavior exists in the current image, determine a category of the suspicious abnormal behavior and a detection frame for the suspicious abnormal behavior. The detection box may cover the area where the suspicious abnormal behavior occurs, and the detection box may be a generally rectangular box. It should be understood that the shape of the detection frame is not limited to a rectangular frame, but may be a circular frame, a square frame, or the like. The abnormal behavior detection model may be constructed using a labeled abnormal behavior dataset with a 2D convolutional neural network, wherein the 2D convolutional neural network is a YOLOV3 target detection network, as shown in fig. 2, wherein the YOLOV3 network is stacked with convolutions of 3x3 and 1x1, does not contain pooling layers, and contracts the picture using convolutional layers.
The abnormal behavior recognition apparatus 100 may further include an abnormal behavior recognition module 120. The abnormal behavior detection module 110 may be further configured to use the detection box to intercept a predetermined number of consecutive images in the video to be identified, starting from the current image. The predetermined number may be 16, and the predetermined number of consecutive images may be truncated at predetermined time intervals. The abnormal behavior recognition module 120 may be configured to acquire a predetermined number of consecutive images and detection frames from the abnormal behavior detection module 110, determine whether there is an abnormal behavior in the consecutive images using an abnormal behavior recognition model, and determine a category of the abnormal behavior when it is determined that there is an abnormal behavior in the consecutive images. The abnormal behavior recognition model 120 may be constructed using a 3D convolutional neural network using training data, wherein the training data is a continuous image that is cut out of the video using the constructed abnormal behavior detection model, and the 3D convolutional neural network may be based on an vgg16 framework, as shown in fig. 3, wherein the 2D convolutional layer and the pooling layer are replaced with a 3D convolutional layer and a 3D pooling layer, and the input layer has dimensions of 16x3x112x 112.
The abnormal behavior recognition apparatus 100 may further include an input module 130 and an output module 140. The input module 110 may be configured to receive a video to be recognized, for example, from a camera, monitor, or the like, and send the video to be recognized to the abnormal behavior detection module 110.
The output module 140 may be configured to obtain and output the category and the detection box of the suspicious abnormal behavior from the abnormal behavior determination module 110, so as to facilitate a security checker, an administrator, and the like to process the suspicious abnormal behavior. In addition, the output module 140 may be further configured to acquire and output the category and the detection frame of the abnormal behavior from the abnormal behavior recognition module 120, so as to facilitate a security inspector, an administrator, and the like to process the abnormal behavior.
According to the embodiment of the disclosure, suspicious abnormal behaviors such as left-over parcels, barrier deliveries, fighting and the like in the video can be found in real time by using the method for detecting the suspicious abnormal behaviors occurring in the images in the video by using the 2D convolutional neural network without selecting an ROI (region of interest) in the video, and the abnormal behaviors such as the left-over parcels, the barrier deliveries, the fighting and the like in the video can be found in real time more accurately by using the method for detecting and identifying the abnormal behaviors occurring in the whole video by using the 2D convolutional neural network and the 3D convolutional neural network.
Fig. 4 shows a flowchart of a method 400 for performing convolutional neural network-based abnormal behavior identification using the convolutional neural network-based abnormal behavior identification apparatus 100 according to an embodiment of the present disclosure.
In step S1, the receiving module 110 may receive a video to be recognized, for example, from a camera, monitor, or the like.
In step S2, the abnormal behavior detection module 110 may acquire the video to be recognized from the reception module 110.
In step S3, the abnormal behavior detection module 110 may detect a current image in the video to be recognized.
In step S4, the abnormal behavior detection module 110 may determine whether there is suspicious abnormal behavior (e.g., a left-over parcel, a barrier delivery, fighting, etc.) in the current image using the abnormal behavior detection model. The abnormal behavior detection model may be constructed using a labeled abnormal behavior dataset with a 2D convolutional neural network, wherein the 2D convolutional neural network is a YOLOV3 target detection network, as shown in fig. 2, wherein the YOLOV3 network is stacked with convolutions of 3x3 and 1x1, does not contain pooling layers, and contracts the picture using convolutional layers.
When it is determined by the abnormal behavior detection module 110 that there is suspicious abnormal behavior in the current image, in step S5, the abnormal behavior detection module 110 may determine suspicious abnormal behavior and a detection frame for the suspicious abnormal behavior. When the abnormal behavior detection module 110 determines that there is no suspicious abnormal behavior in the current image, the current flow ends.
After step S5, the output module 140 may acquire and output the category and the detection box of the suspicious abnormal behavior from the abnormal behavior detection module 110, so that a security inspector, an administrator, or the like can handle the suspicious abnormal behavior.
Fig. 5 shows a flowchart of another abnormal behavior recognition method 500 based on a convolutional neural network using the abnormal behavior recognition apparatus 100 based on a convolutional neural network according to an embodiment of the present disclosure. Steps S1 to S5 in the abnormal behavior recognition method 500 are the same as steps S1 to S5 in the abnormal behavior recognition method 400, and thus are not described herein again.
In step S6, the abnormal behavior detection module 110 may intercept a predetermined number of consecutive images in the video to be recognized from the current image using a detection box, where the detection box may cover an area where the suspicious abnormal behavior occurs, and the detection box may be a rectangular box. It should be understood that the shape of the detection frame is not limited to a rectangular frame, but may be a circular frame, a square frame, or the like. The predetermined number may be 16, and the predetermined number of consecutive images may be truncated at predetermined time intervals.
In step S7, the abnormal behavior recognition module 120 may acquire a predetermined number of consecutive images and detection frames from the abnormal behavior detection module 110.
In step S8, the abnormal behavior recognition module 120 may determine whether there is abnormal behavior in the continuous image using the abnormal behavior recognition model. The abnormal behavior recognition model may be constructed using a 3D convolutional neural network using training data, wherein the training data is a continuous image that is cut out of the video using the constructed abnormal behavior detection model, and the 3D convolutional neural network may be based on an vgg16 framework, as shown in fig. 3, wherein the 2D convolutional layer and the pooling layer are replaced with a 3D convolutional layer and a 3D pooling layer, and the input layer has dimensions of 16x3x112x 112.
When the abnormal behavior recognition module determines that there is abnormal behavior in the continuous images, the abnormal behavior recognition module 120 may determine the category of the abnormal behavior in step S9. When the abnormal behavior recognition module 120 determines that there is no abnormal behavior in the continuous images, the current flow ends.
After step S9, the output module 140 may acquire and output the category and the detection frame of the abnormal behavior from the abnormal behavior recognition module 120, so that a security inspector, an administrator, or the like can handle the abnormal behavior.
According to the embodiment of the disclosure, the abnormal behavior action detection of the single-frame video image based on the 2D convolutional neural network and the action identification method of the intercepted continuous multi-frame video image based on the 3D convolutional neural network can accurately position the position where the abnormal behavior occurs in the whole scene, and can also identify the occurring abnormal behavior in real time and efficiently. In the method, firstly, a 2D convolutional neural network is used for detecting suspicious abnormal behavior actions in a video image and obtaining a detection frame; then, a plurality of continuous images are intercepted by using a detection frame, and the continuous images are sent to a 3D convolutional neural network to identify abnormal behaviors such as left-over parcels, barrier deliveries, fighting and the like. By the method, abnormal behaviors occurring in the video can be quickly detected and identified without selecting the ROI, so that the omnibearing monitoring is realized.
Fig. 6 shows a schematic diagram of a convolutional neural network-based abnormal behavior identification system, according to an embodiment of the present disclosure. The system 600 may include a processor 610, such as a Digital Signal Processor (DSP). Processor 610 may be a single device or multiple devices for performing different acts of the processes described herein. The system 600 may also include input/output (I/O) devices 630 to receive signals from or transmit signals to other entities.
Further, the system 600 may include a memory 620, the memory 620 may be of the form: non-volatile or volatile memory, such as electrically erasable programmable read-only memory (EEPROM), flash memory, and the like. Memory 620 may store computer readable instructions that, when executed by processor 610, may cause the processor to perform the actions described herein.
Some block diagrams and/or flow diagrams are shown in the figures. It will be understood that some blocks of the block diagrams and/or flowchart illustrations, or combinations thereof, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the instructions, which execute via the processor, create means for implementing the functions/acts specified in the block diagrams and/or flowchart block or blocks.
Accordingly, the techniques of this disclosure may be implemented in hardware and/or software (including firmware, microcode, etc.). In addition, the techniques of this disclosure may take the form of a computer program product on a computer-readable medium having instructions stored thereon for use by or in connection with an instruction execution system (e.g., one or more processors). In the context of this disclosure, a computer-readable medium may be any medium that can contain, store, communicate, propagate, or transport the instructions. For example, the computer readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, or propagation medium. Specific examples of the computer readable medium include: magnetic storage devices, such as magnetic tape or Hard Disk Drives (HDDs); optical storage devices, such as compact disks (CD-ROMs); a memory, such as a Random Access Memory (RAM) or a flash memory; and/or wired/wireless communication links.
The foregoing detailed description has set forth numerous embodiments of convolutional neural network-based abnormal behavior identification methods, apparatus, and systems using schematics, flowcharts, and/or examples. Where such diagrams, flowcharts, and/or examples contain one or more functions and/or operations, it will be understood by those within the art that each function and/or operation within such diagrams, flowcharts, or examples can be implemented, individually and/or collectively, by a wide range of structures, hardware, software, firmware, or virtually any combination thereof. In one embodiment, portions of the subject matter described in embodiments of the present disclosure may be implemented by Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs), Digital Signal Processors (DSPs), or other integrated formats. However, those skilled in the art will recognize that some aspects of the embodiments disclosed herein, in whole or in part, can be equivalently implemented in integrated circuits, as one or more computer programs running on one or more computers (e.g., as one or more programs running on one or more computer systems), as one or more programs running on one or more processors (e.g., as one or more programs running on one or more microprocessors), as firmware, or as virtually any combination thereof, and that designing the circuitry and/or writing the code for the software and or firmware would be well within the skill of one of skill in the art in light of this disclosure. In addition, those skilled in the art will appreciate that the mechanisms of the subject matter described herein are capable of being distributed as a program product in a variety of forms, and that an illustrative embodiment of the subject matter described herein applies regardless of the particular type of signal bearing media used to actually carry out the distribution. Examples of signal bearing media include, but are not limited to: recordable type media such as floppy disks, hard disk drives, Compact Disks (CDs), Digital Versatile Disks (DVDs), digital tape, computer memory, etc.; and a transmission type medium such as a digital and/or an analog communication medium (e.g., a fiber optic cable, a waveguide, a wired communications link, a wireless communication link, etc.).
Claims (10)
1. An abnormal behavior recognition apparatus based on a convolutional neural network, comprising:
the abnormal behavior detection module is configured to acquire a video to be identified, detect a current image in the video to be identified, determine whether suspicious abnormal behaviors exist in the current image by using an abnormal behavior detection model, and when the suspicious abnormal behaviors exist in the current image, determine the category of the suspicious abnormal behaviors and a detection frame aiming at the suspicious abnormal behaviors, wherein the detection frame covers an area where the suspicious abnormal behaviors occur.
2. The abnormal behavior recognition device according to claim 1,
the abnormal behavior detection module is further configured to intercept a predetermined number of continuous images in the video to be identified from the current image using the detection frame; and is
The abnormal behavior recognizing apparatus further includes: an abnormal behavior recognition module configured to acquire the predetermined number of consecutive images and the detection frame from the abnormal behavior detection module, determine whether an abnormal behavior exists in the consecutive images using an abnormal behavior recognition model, and determine a category of the abnormal behavior when it is determined that the abnormal behavior exists in the consecutive images.
3. The abnormal behavior recognition device according to claim 2, further comprising:
a receiving module configured to receive the video to be identified and send the video to be identified to the abnormal behavior detection module; and
an output module configured to acquire and output the category of the suspicious abnormal behavior and the detection box from the abnormal behavior determination module.
4. The abnormal behavior recognition apparatus according to claim 3, wherein the output module is further configured to acquire and output the category of the abnormal behavior and the detection box from the abnormal behavior recognition module.
5. The abnormal behavior recognition apparatus of claim 1, wherein the abnormal behavior detection model is constructed using an abnormal behavior data set with a 2D convolutional neural network.
6. The abnormal behavior recognition apparatus according to claim 2, wherein the abnormal behavior recognition model is constructed using a 3D convolutional neural network using training data which is a continuous image cut out from a video using the constructed abnormal behavior detection model.
7. An abnormal behavior identification method based on a convolutional neural network comprises the following steps:
acquiring a video to be identified by an abnormal behavior detection module;
detecting a current image in the video to be identified by the abnormal behavior detection module;
determining, by the abnormal behavior detection module, whether there is suspicious abnormal behavior in the current image using an abnormal behavior detection model; and
when the abnormal behavior detection module determines that suspicious abnormal behaviors exist in the current image, the abnormal behavior detection module determines the class of the suspicious abnormal behaviors and a detection frame aiming at the suspicious abnormal behaviors, wherein the detection frame covers the area where the suspicious abnormal behaviors occur.
8. The abnormal behavior recognition method of claim 7, further comprising:
intercepting, by the abnormal behavior detection module, a predetermined number of consecutive images in the video to be recognized starting from the current image using the detection frame;
acquiring, by an abnormal behavior recognition module, the predetermined number of consecutive images and the detection frame from the abnormal behavior detection module;
determining, by the abnormal behavior recognition module, whether there is an abnormal behavior in the continuous image using an abnormal behavior recognition model; and
when the abnormal behavior identification module determines that abnormal behavior exists in the continuous images, determining the category of the abnormal behavior by the abnormal behavior identification module.
9. The abnormal behavior recognition method of claim 8, further comprising:
receiving the video to be identified by a receiving module and sending the video to be identified to the abnormal behavior detection module; and
obtaining and outputting, by an output module, the category of the suspicious abnormal behavior and the detection box from the abnormal behavior determination module.
10. The abnormal behavior recognition method of claim 9, further comprising:
and acquiring and outputting the category of the abnormal behavior and the detection frame from the abnormal behavior identification module by the output module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911262520.9A CN112949359A (en) | 2019-12-10 | 2019-12-10 | Convolutional neural network-based abnormal behavior identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911262520.9A CN112949359A (en) | 2019-12-10 | 2019-12-10 | Convolutional neural network-based abnormal behavior identification method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112949359A true CN112949359A (en) | 2021-06-11 |
Family
ID=76226005
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911262520.9A Pending CN112949359A (en) | 2019-12-10 | 2019-12-10 | Convolutional neural network-based abnormal behavior identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112949359A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113269500A (en) * | 2021-06-16 | 2021-08-17 | 江苏佳利达国际物流股份有限公司 | Cold-chain logistics monitoring method and system based on neural network |
| CN113989608A (en) * | 2021-12-01 | 2022-01-28 | 西安电子科技大学 | Student experiment classroom behavior identification method based on top vision |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107766839A (en) * | 2017-11-09 | 2018-03-06 | 清华大学 | Action identification method and device based on neutral net |
| CN109241946A (en) * | 2018-10-11 | 2019-01-18 | 平安科技（深圳）有限公司 | Abnormal behaviour monitoring method, device, computer equipment and storage medium |
| CN109508638A (en) * | 2018-10-11 | 2019-03-22 | 平安科技（深圳）有限公司 | Face Emotion identification method, apparatus, computer equipment and storage medium |
| CN109635790A (en) * | 2019-01-28 | 2019-04-16 | 杭州电子科技大学 | A kind of pedestrian's abnormal behaviour recognition methods based on 3D convolution |
| CN110189378A (en) * | 2019-05-23 | 2019-08-30 | 北京奇艺世纪科技有限公司 | A kind of method for processing video frequency, device and electronic equipment |
| WO2019184124A1 (en) * | 2018-03-30 | 2019-10-03 | 平安科技（深圳）有限公司 | Risk-control model training method, risk identification method and apparatus, and device and medium |
| CN110414313A (en) * | 2019-06-06 | 2019-11-05 | 平安科技（深圳）有限公司 | Abnormal behaviour alarm method, device, server and storage medium |
-
2019
- 2019-12-10 CN CN201911262520.9A patent/CN112949359A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107766839A (en) * | 2017-11-09 | 2018-03-06 | 清华大学 | Action identification method and device based on neutral net |
| WO2019091417A1 (en) * | 2017-11-09 | 2019-05-16 | 清华大学 | Neural network-based identification method and device |
| WO2019184124A1 (en) * | 2018-03-30 | 2019-10-03 | 平安科技（深圳）有限公司 | Risk-control model training method, risk identification method and apparatus, and device and medium |
| CN109241946A (en) * | 2018-10-11 | 2019-01-18 | 平安科技（深圳）有限公司 | Abnormal behaviour monitoring method, device, computer equipment and storage medium |
| CN109508638A (en) * | 2018-10-11 | 2019-03-22 | 平安科技（深圳）有限公司 | Face Emotion identification method, apparatus, computer equipment and storage medium |
| CN109635790A (en) * | 2019-01-28 | 2019-04-16 | 杭州电子科技大学 | A kind of pedestrian's abnormal behaviour recognition methods based on 3D convolution |
| CN110189378A (en) * | 2019-05-23 | 2019-08-30 | 北京奇艺世纪科技有限公司 | A kind of method for processing video frequency, device and electronic equipment |
| CN110414313A (en) * | 2019-06-06 | 2019-11-05 | 平安科技（深圳）有限公司 | Abnormal behaviour alarm method, device, server and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 沈铮等: "基于图像处理的公交车内人群异常情况检测", 《计算机工程与设计》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113269500A (en) * | 2021-06-16 | 2021-08-17 | 江苏佳利达国际物流股份有限公司 | Cold-chain logistics monitoring method and system based on neural network |
| CN113989608A (en) * | 2021-12-01 | 2022-01-28 | 西安电子科技大学 | Student experiment classroom behavior identification method based on top vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200184103A1 (en) | Methods and apparatus for using video analytics to detect regions for privacy protection within images from moving cameras | |
| AU2017436901B2 (en) | Methods and apparatus for automated surveillance systems | |
| JP7229662B2 (en) | How to issue alerts in a video surveillance system | |
| CN108228705B (en) | Automatic object and activity tracking device, method and medium in live video feedback | |
| US8743204B2 (en) | Detecting and monitoring event occurrences using fiber optic sensors | |
| CN111680535B (en) | Method and system for real-time prediction of one or more potential threats in video surveillance | |
| CN101208710A (en) | Target detection and tracking from overhead video streams | |
| US10997469B2 (en) | Method and system for facilitating improved training of a supervised machine learning process | |
| CN107122743B (en) | Security monitoring method and device and electronic equipment | |
| CN109544870B (en) | Alarm judgment method for intelligent monitoring system and intelligent monitoring system | |
| CN112949359A (en) | Convolutional neural network-based abnormal behavior identification method and device | |
| KR101454644B1 (en) | Loitering Detection Using a Pedestrian Tracker | |
| CN111401239B (en) | Video analysis method, device, system, equipment and storage medium | |
| CN114663871A (en) | Image recognition method, training method, device, system and storage medium | |
| CN113158963A (en) | High-altitude parabolic detection method and device | |
| KR102584708B1 (en) | System and Method for Crowd Risk Management by Supporting Under and Over Crowded Environments | |
| US20230215128A1 (en) | Real-time occlusion removal using synthetic pixel generation | |
| CN114842411A (en) | Group behavior identification method based on complementary space-time information modeling | |
| Xie et al. | On‐line physical security monitoring of power substations | |
| CN114782883A (en) | Abnormal behavior detection method, device and equipment based on group intelligence | |
| US11227162B1 (en) | Multilayer information dynamics for activity and behavior detection | |
| US11756217B2 (en) | Multi-camera system to perform movement pattern anomaly detection | |
| KR102641828B1 (en) | Providing method, apparatus and computer-readable medium of recognition of military dangerous behaviors in videos and tracking of dangerous behaviors | |
| US20240046647A1 (en) | Method and device for detecting obstacles, and computer storage medium | |
| CN117437529A (en) | Cross-attention mechanism-based target detection model construction method and application |
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 |