Behavior monitoring method, and electronic device and computer-readable storage medium
By privacy processing of abnormal behavioral data monitored by image sensors and generating behavioral events, the problem of privacy and security risks in the driving monitoring system is solved, and low-cost and privacy behavioral monitoring is achieved to adapt to market demand.
Patent Information
- Application Number
- PCT/CN2023/111791
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-10
- Filing Date
- 2023-08-08
- Publication Date
- 2024-11-14
AI Technical Summary
The existing driving monitoring system has privacy and security risks when monitoring driver behavior, and has high cost and poor universality, which cannot meet market demand.
By evaluating abnormal behaviors and privacy processing of relevant data when the image sensor monitors personnel behavior, generating personnel behavior events to ensure privacy and security, while continuing to use low-cost image sensors for monitoring to avoid hardware configuration adjustments.
It effectively avoids the leakage of real images of people during behavior monitoring, ensures privacy and security, and reduces monitoring costs and adapts to market demand.
Smart Images

Figure CN2023111791_14112024_PF_FP_ABST
Abstract
Description
Method, electronic device and computer-readable storage medium for behavior monitoring
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS
[0002] This disclosure claims priority to U.S. Provisional Application No. 63 / 465,262, filed on May 10, 2023, entitled “Monitor system.” Technical Field
[0003] The present disclosure generally relates to the field of image processing technology. More specifically, the present disclosure relates to a method for behavior monitoring, and an electronic device and a computer-readable storage medium for executing the method. Background Art
[0004] With the rapid development of information technology, while it brings convenience to people, it also faces the problem of privacy and security. For example, in the application scenario of the Driver Monitor System (DMS), the driver is usually monitored for driving behaviors such as playing with mobile phones and driving while fatigued. Once the driver is monitored to have the above driving behaviors during driving, the image of the driver will be output. However, the driver's image will directly show the true state of the driver's facial features. Although it realizes the monitoring of driving behavior, it also leads to privacy and security risks in the monitoring process of personnel behavior.
[0005] However, current monitoring systems with privacy processing (for example, systems equipped with dynamic vision sensors, or DVS) have the disadvantages of high cost and poor universality, and cannot meet market demand.
[0006] Summary of the Invention
[0007] In order to at least solve the technical problems described in the above background technology section, the present disclosure proposes a method, electronic device and computer-readable storage medium for behavior monitoring to solve the problem of privacy and security risks in personnel behavior monitoring in the existing technology, and can ensure the privacy and security of personnel behavior monitoring during the process of behavior monitoring.
[0008] In a first aspect, the present disclosure provides a method for behavior monitoring, comprising: in the process of monitoring personnel behavior based on image sensors, evaluating personnel behavior based on monitoring data of the image sensors; in response to an evaluation result that there is abnormal behavior, obtaining unprocessed data associated with the abnormal behavior; performing privacy processing on the unprocessed data; and generating personnel behavior events based on the privacy-processed data and the evaluation result of the abnormal behavior.
[0009] In one embodiment, obtaining the data to be processed associated with the abnormal behavior includes: determining the time of occurrence of the abnormal behavior, and using the monitoring data starting from the time of occurrence of the abnormal behavior as the data to be processed; and / or determining the evaluation characteristics of the abnormal behavior, and using the monitoring data containing the evaluation characteristics as the data to be processed.
[0010] In another embodiment, performing privacy processing on the data to be processed includes: performing blurring processing on at least a portion of the data to be processed.
[0011] In yet another embodiment, performing fuzzy processing on at least a partial area of the data to be processed includes: performing fuzzy processing on the entire image of the data to be processed.
[0012] In yet another embodiment, performing fuzzy processing on at least a portion of the area in the data to be processed includes: performing fuzzy processing on a human body area in the data to be processed.
[0013] In yet another embodiment, performing fuzzy processing on at least a portion of the area in the data to be processed includes: performing fuzzy processing on an area where a designated feature in the data to be processed is located.
[0014] In yet another embodiment, the designated features include features of a person's head and face and / or limbs.
[0015] In another embodiment, fuzzifying at least a portion of the area in the data to be processed includes: obtaining evaluation features of the abnormal behavior; determining the location area of the evaluation features in the data to be processed; and fuzzifying other areas in the data to be processed except the location area.
[0016] In another embodiment, blurring the area where the designated feature in the data to be processed is located includes: performing graded blurring on the designated feature according to the privacy level of the designated feature; wherein the privacy level of the designated feature is positively correlated with the level of blurring.
[0017] In another embodiment, the data to be processed is blurred by any of the following operations: performing differentiation processing on the front and back pictures in the data to be processed; performing edge difference processing on the data to be processed; reducing the resolution of the data to be processed; reducing the size of the data to be processed; performing Canny blur processing, Gaussian blur processing and / or average value blur processing on the data to be processed.
[0018] In another embodiment, generating the personnel behavior event includes: fusing the privacy-processed data with the evaluation result of the abnormal behavior to obtain the personnel behavior event.
[0019] In yet another embodiment, the method further includes: displaying the personnel behavior event; and / or storing the personnel behavior event locally or in the cloud.
[0020] In a second aspect, the present disclosure provides an electronic device comprising: a processor; and a memory storing computer instructions for behavior monitoring, wherein when the computer instructions are executed by the processor, the electronic device executes the method according to the above and following embodiments.
[0021] In a third aspect, the present disclosure provides a computer-readable storage medium having a computer program stored therein, wherein the computer program is configured to execute any one of the aforementioned methods for behavior monitoring when run.
[0022] Different from the existing technology, the method for behavior monitoring provided by the present disclosure, in the process of monitoring personnel behavior using image sensors, performs privacy processing on the unprocessed data related to abnormal behavior, and combines it with the evaluation results of abnormal behavior to obtain personnel behavior events. It can be seen that in the process of personnel behavior monitoring, the solution of the present disclosure can, on the one hand, use personnel behavior events to record personnel behavior, thereby effectively avoiding the leakage of real images of personnel, thereby ensuring privacy security in the process of behavior monitoring. On the other hand, low-cost image sensors can continue to be used to implement personnel behavior monitoring, that is, the implementation of the entire solution does not require excessive adjustment of hardware configuration, thereby achieving low-cost and private behavior monitoring.
[0023] The computer-readable storage medium and electronic device provided by the present disclosure belong to the same inventive concept as the method for behavior monitoring, and therefore have the same beneficial effects, which will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] By reading the following detailed description with reference to the accompanying drawings, the description of the exemplary embodiments of the present disclosure and other objects, features and advantages will become easier to understand. In the accompanying drawings, several embodiments of the present disclosure are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:
[0025] FIG1 is a flow chart of a method for behavior monitoring provided by a first embodiment of the present disclosure;
[0026] FIG2 is a flow chart of a method for behavior monitoring provided by a second embodiment of the present disclosure;
[0027] FIG3 is a flow chart of a method for behavior monitoring provided by a third embodiment of the present disclosure; and
[0028] FIG4 is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION
[0029] The following is a more detailed description of the specific embodiments of the present disclosure, with reference to schematic diagrams. The advantages and features of the present disclosure will become more apparent from the following description and claims. It should be noted that the drawings are in a very simplified form and are not to exact scale, and are only used to facilitate and clearly illustrate the purpose of the embodiments of the present disclosure.
[0030] In the description of the present disclosure, it should be understood that the terms "center", "up", "down", "left", "right", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present disclosure and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as a limitation on the present disclosure.
[0031] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. Throughout the present disclosure, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.
[0032] The specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
[0033] FIG1 schematically shows a diagram of a method 100 for behavior monitoring according to an embodiment of the present disclosure.
[0034] It should be noted that the method 100 in this embodiment can be applied to any application scenario (such as a DMS scenario, etc.) that requires behavior monitoring with the help of images.
[0035] As shown in Figure 1, at step S101, in the process of monitoring personnel behavior based on image sensors, personnel behavior can be evaluated based on the monitoring data of the image sensor. It should be noted that the specific model and installation location of the image sensor are not limited here, and can be set and adjusted in combination with the actual application scenario. For example, in the DMS scenario, the image sensor is installed inside the vehicle, and personnel behavior can be monitored in real time based on the image sensor. It can be understood that the image sensor here can include an RGB sensor, or other image sensors with privacy processing requirements. Preferably, the image sensor involved in the present disclosure can adopt a low-cost image sensor that can support personnel behavior monitoring.
[0036] In some embodiments, the monitoring data based on the image sensor can be evaluated in combination with a pre-trained behavior evaluation model (such as an AI model, etc.) to evaluate whether there are abnormal behaviors (such as dozing off, yawning, playing with mobile phones, conflicts with others, etc.). It should be noted that the specific behavior types of abnormal behaviors are not limited here, and can be set according to the application scenario and actual design requirements. For example, in a driving application scenario, abnormal behaviors may include some behaviors that may affect driving safety (such as fatigue driving, answering and making phone calls, conflicts with passengers, etc.). For another example, for application scenarios with high environmental quality requirements, abnormal behaviors may include some behaviors that affect environmental quality (such as smoking, etc.).
[0037] In addition, it should be noted that the aforementioned use of image sensors for behavior monitoring and personnel behavior assessment can be implemented using traditional technologies (such as training AI models to assess personnel behavior, triggering image sensors to monitor personnel behavior, etc.), so the relevant details will not be repeated here. Of course, it is also possible to combine specific application scenarios to target the selection of AI model training data and AI model base model, as well as the trigger conditions for triggering the image sensor to start working and the specific format of the monitoring data.
[0038] At step S102, in response to the assessment result indicating the presence of abnormal behavior, data to be processed associated with the abnormal behavior is obtained. During the process of assessing human behavior based on monitoring data from an image sensor, the monitoring data can be used to determine whether the human behavior is normal or abnormal. When the assessment result indicates the presence of abnormal behavior, data to be processed associated with the abnormal behavior can be obtained. In some embodiments, the aforementioned monitoring data may include image data, and when the presence of abnormal behavior is determined, the data to be processed can be obtained based on the monitoring data.
[0039] Next, in step S103, the aforementioned data to be processed may be privacy-enhanced. When the data to be processed is obtained, it may be privacy-enhanced, particularly data in the processed data that may leak privacy (e.g., information about the person's own body in the image, information about the person's environment, information about the person's clothing, etc.).
[0040] Finally, at step S104, a human behavior event can be generated based on the privacy-processed data and the abnormal behavior assessment results. After the privacy-processed data to be processed is completed, the privacy-processed data can be obtained. Furthermore, the privacy-processed data and the abnormal behavior assessment results can be used to generate human behavior events.
[0041] Therefore, during the entire personnel behavior monitoring process, personnel behavior can be monitored based on the personnel behavior events obtained after privacy processing, thereby effectively avoiding the leakage of some personnel privacy and sensitive information, thereby ensuring privacy security during the behavior monitoring process. In addition, during the specific implementation of the solution disclosed in the present invention, low-cost image sensors (such as RGB image sensors) can continue to be used to implement personnel behavior monitoring, that is, the implementation of the entire solution does not require excessive adjustment of the hardware configuration, thereby achieving low-cost and privacy-oriented behavior monitoring. In addition, during the specific implementation of the solution disclosed in the present invention, traditional evaluation models can also be used to implement behavior evaluation, thereby effectively shortening the research and development cycle of the entire solution, making the solution disclosed in the present invention have strong compatibility and universality, and more adapted to market demand.
[0042] FIG2 schematically shows a diagram of a method 200 for behavior monitoring according to another embodiment of the present disclosure.
[0043] It is understandable that method 200 is a further limitation and / or extension of method 100 in Figure 1. Therefore, the above detailed description in conjunction with Figure 1 is also applicable to the following.
[0044] As shown in FIG2 , at step S201, during the process of monitoring human behavior using an image sensor, human behavior can be evaluated based on the monitoring data from the image sensor. As previously described, human behavior monitoring can be implemented in any application scenario requiring behavioral regulation or monitoring using any image sensor with privacy protection requirements (e.g., an RGB image sensor). During the process of monitoring human behavior using an image sensor, human behavior can be evaluated based on the monitoring data from the image sensor using a pre-trained behavior evaluation model.
[0045] Then, when it is determined that there is an evaluation result of abnormal behavior, the data to be processed associated with the abnormal behavior can be obtained. Specifically, at step S202, the data to be processed can be determined based on the time of occurrence of the abnormal behavior and / or the evaluation characteristics of the abnormal behavior. Specifically, in some embodiments, the time of occurrence of the abnormal behavior can be determined, and the monitoring data starting from the time of occurrence of the abnormal behavior can be used as the data to be processed. For example, the monitoring data within a period of time starting from the time of occurrence of the abnormal behavior can be used as the data to be processed. The monitoring data can be a video or a few frames of pictures within the time interval. Among them, the aforementioned time interval can be set or adjusted according to needs.
[0046] In other embodiments, evaluation features of abnormal behavior can also be determined, and monitoring data containing such evaluation features can be used as data to be processed. The evaluation features here can be understood as features used by the behavior assessment model to determine abnormal behavior during the behavior assessment process (e.g., mouth features, eye features, hand features, etc.). When monitoring human behavior, monitoring data containing such evaluation features can be used as data to be processed.
[0047] In yet other embodiments, the data to be processed can be determined by combining the abnormal behavior's occurrence time and evaluation characteristics. That is, data obtained based on either the abnormal behavior's occurrence time or evaluation characteristics can be further filtered using the other condition to obtain the data to be processed. Thus, by using the abnormal behavior's occurrence time and / or evaluation characteristics to obtain data highly relevant to the abnormal behavior, interference from irrelevant data can be avoided, thereby improving the accuracy and efficiency of subsequent privacy processing.
[0048] Next, the aforementioned acquired data to be processed may be privacy-processed.
[0049] Specifically, at step S203, at least a part of the area in the data to be processed can be blurred. For example, in some embodiments, the data to be processed can be blurred as a whole. In this embodiment, the data to be processed may include videos or pictures, etc., and the whole picture can be blurred based on at least one frame of the video or picture. For example, the blurring of the whole picture can be achieved by using technologies such as difference processing technology between different pictures, edge difference processing technology, resolution reduction technology, size reduction, Canny blur processing, Gaussian blur processing and / or average value blur. Of course, in addition to the above-mentioned blurring processing methods, blurring can also be achieved by degrading the color level details of the picture and causing the color block to be disrupted, or by using pictures to construct a virtual portrait. It should be noted that the detailed description of the blurring processing method here is only an exemplary description, and the solution disclosed in the present invention is not limited to this. It can be adjusted in combination with application requirements.
[0050] In other embodiments, it is also possible to perform fuzzy processing on some areas in the data to be processed. For example, the human body area in the data to be processed can be fuzzy processed, that is, the background area is retained and only the human body area is fuzzy processed. For another example, it is also possible to perform fuzzy processing on the area where the specified features in the data to be processed are located. Specifically, in some implementation scenarios, it is possible to perform fuzzy processing on the specified features such as the head and face features and / or limbs of the person, or to perform fuzzy processing on the specified features such as sensitive decorations worn by the person (such as clothes, headwear, earrings, rings, watches, etc.), sensitive information in the background (such as car logos, location signs, etc.). It is understandable that in this scenario, the aforementioned specified features can be customized, and can be set and adjusted specifically according to the application requirements.
[0051] In other implementation scenarios, when fuzzifying some areas in the data to be processed, it is possible to obtain the evaluation features of abnormal behavior (such as mouth features, eye features, etc.), and then determine the location area of the evaluation features of abnormal behavior in the data to be processed, and retain the location area, while fuzzifying other areas in the data to be processed except for the location area. For example, when a person has abnormal behaviors such as turning his head or closing his eyes, and the abnormal behavior is detected based on the eye features, the location area of the eyes in the data to be processed can be retained, and the other areas except for the area where the eyes are located can be fuzzified. In this way, other areas outside the area where the eyes are located will not be truly displayed later, which improves privacy and security while meeting the needs of personal behavior supervision.
[0052] Furthermore, in the process of fuzzifying the area where the designated features in the data to be processed are located, the designated features can be graded and fuzzified according to the privacy level of the designated features. For example, the privacy level of each designated feature can be pre-set, as well as the correlation between the privacy level and the different levels of fuzzification. Among them, the higher the privacy level of the designated feature, the higher the privacy level. At this time, the privacy level of the designated feature is positively correlated with the level of fuzzification. That is, different levels of fuzzification need to be performed on designated features with different levels of privacy. For example, a designated feature with a low level of privacy can be subjected to a slight degree of fuzzification, while a designated feature with a high level of privacy can be subjected to the highest degree of fuzzification. It should be noted that the description of the hierarchical fuzzification process here is only an exemplary description.
[0053] In addition, when blurring partial areas of the processed data, blurring can be achieved by performing image difference processing, edge difference processing, reducing resolution, reducing size, Canny blur processing, Gaussian blur processing and / or average value blurring on partial areas of different images, degrading color level details in the image and causing a color block disruption effect, or using images to construct virtual portraits.
[0054] After the privacy-enhanced data is processed, at step S204, the privacy-enhanced data can be fused with the abnormal behavior assessment results to obtain the personnel behavior events. This fusion process can be performed by labeling, overlaying, or integrating the assessment results and the privacy-enhanced data using traditional image fusion technology.
[0055] Finally, at step S205, the aforementioned human behavior events can be displayed and / or stored locally or in the cloud. Human behavior events are obtained by privacy-processing data and evaluation results. They can be displayed on the vehicle's internal display, displayed via an external display, or stored locally or in the cloud, effectively preventing privacy leaks.
[0056] Therefore, privacy protection can be achieved by obtaining the data to be processed and blurring at least some of its regions. Furthermore, by integrating the privacy-protected data with the abnormal behavior assessment results, we can generate behavioral events that can be used to record abnormal behavior. This behavior monitoring process effectively reduces the risk of privacy leaks.
[0057] FIG3 schematically shows a diagram of a method 300 for behavior monitoring according to yet another embodiment of the present disclosure.
[0058] It is understandable that method 300 is a specific technical implementation of method 100 and method 200. Therefore, the above detailed description in conjunction with FIG1 and FIG2 is also applicable to the following.
[0059] This embodiment illustrates the specific application of method 100 or method 200 in a DMS scenario. In this DMS scenario, the vehicle may be equipped with an RGB image sensor, which supports driver behavior monitoring and driver behavior assessment based on the RGB image sensor. Therefore, this embodiment can achieve driver behavior monitoring by further optimizing the monitoring process based on the combination of low-cost RGB image sensors and existing driver behavior assessment technology.
[0060] As shown in FIG3 , at step S301, the driver's behavior can be evaluated based on the monitoring data of the RGB image sensor. Specifically, during the driving process, the driver's behavior can be monitored in real time by the RGB image sensor, and combined with a pre-trained artificial intelligence behavior evaluation model to evaluate whether the driver has abnormal behaviors such as fatigue driving, making phone calls, conflicts with passengers, and other dangerous driving. It should be noted that the behavior evaluation model here can be a pre-trained evaluation model that can be used directly. Of course, in actual applications, the training samples and specific optimization strategies can be adjusted according to the application scenario to retrain a behavior evaluation model that supports various behavior evaluations.
[0061] At step S302, in response to the evaluation result of the existence of abnormal behavior, the monitoring data starting from the time of occurrence of the abnormal behavior can be used as data to be processed. Specifically, when the existence of abnormal behavior is evaluated using the trained behavior evaluation model, the monitoring data within a period of time (the length of the time interval can be adjusted according to application requirements) starting from the time of occurrence of the abnormal behavior are all used as data to be processed (such as a video or several frames of pictures). The description of the process of obtaining the data to be processed here is only an exemplary description, and the scheme of the present disclosure is not limited to this. For example, the evaluation features based on which the abnormal behavior evaluation process is based can also be determined to determine the data to be processed. Specifically, monitoring data containing evaluation features (such as whether the mouth is open, whether the eyes are open, hand posture, etc.) can be used as data to be processed.
[0062] Next, the data to be processed can be privacy-enhanced. Specifically, at step S303, the difference between the two pictures before and after in the data to be processed can be used to implement privacy-enhanced processing. For example, the difference between the two frames of the data to be processed can be compared, and privacy-enhanced processing can be implemented based on the difference results. For another example, the privacy-enhancing processing of the image can be implemented by calculating the difference between two frames of images with a fixed interval (for example, 33ms) of difference. For another example, the privacy-enhancing processing of the image can be implemented by edge difference processing of the image, reducing the image resolution, reducing the image size, rasterizing the image, Canny blur processing, Gaussian blur processing, average blur, image mosaic, constructing a virtual portrait, etc. It should be noted that the description of the privacy-enhancing processing process here is only an example description, and the appropriate processing means can be selected according to the application requirements to implement the privacy-enhancing processing of the image.
[0063] Finally, at step S304, the privacy-decorated data and the abnormal behavior assessment results can be combined to obtain the driver behavior event. The privacy-decorated data may include a blurred image. In this scenario, the assessment results can be integrated with the blurred image to obtain the driver behavior event.
[0064] Therefore, the disclosed solution can use traditional RGB sensors during behavior monitoring, avoiding excessive hardware adjustments and effectively controlling the implementation costs of the technology. Furthermore, the entire solution can be implemented in conjunction with current behavior assessment technologies, eliminating the need for additional training of relevant behavior assessment models, allowing the entire technical solution to be implemented more quickly.
[0065] It should be noted that the solution of the present disclosure is described here using the DMS scenario as an example. In practical applications, methods 100 and 200 are not limited to the DMS scenario, but can be applied to any application scenario that requires the use of images to implement behavior monitoring. Among them, in various application scenarios, low-cost RGB sensors can be used to implement personnel behavior monitoring. Of course, other low-cost, highly circulated image collectors that also have privacy and security issues can also be used to replace the RGB sensor.
[0066] FIG4 schematically shows a schematic block diagram of an electronic device 400 according to an embodiment of the present disclosure.
[0067] As shown in Figure 4 , electronic device 400 may include a processor 401 and a memory 402. Figure 4 also shows an image sensor 403, which can be configured to monitor human behavior and provide monitoring data regarding that behavior. In practical applications, image sensor 403 may be integrated into electronic device 400, or may be provided separately from electronic device 400. It should be noted that the specific type of image sensor is not limited here, and may include an RGB sensor or other image sensor with privacy protection requirements.
[0068] Memory 402 stores computer instructions for behavior monitoring. When executed by processor 401, the computer instructions cause electronic device 400 to perform the method described above in conjunction with Figures 1 to 3. For example, in some embodiments, electronic device 400 can utilize monitoring data from an RGB image sensor to assess human behavior, assess abnormal behavior, obtain unprocessed data associated with abnormal behavior, perform privacy-enhancing processing on unprocessed data, generate human behavior events, and perform hierarchical fuzzy processing on designated features. Based on this, electronic device 400 can, during the human behavior monitoring process, utilize human behavior events to record human behavior, thereby effectively preventing the leakage of real images of the person, thereby ensuring privacy security during the behavior monitoring process. On the other hand, low-cost RGB image sensors can continue to be used to implement human behavior monitoring. In other words, the implementation of the entire solution does not require excessive adjustments to the hardware configuration, thereby achieving low-cost and private behavior monitoring.
[0069] In addition, in a specific application, the method described in FIG. 1 to FIG. 3 can be implemented by a system on chip (System on Chip, SOC for short) or an application specific integrated circuit (Application Specific Integrated Circuit, ASIC for short).
[0070] It should be noted that although the detailed description above mentions several devices or sub-devices of a device, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more devices described above can be embodied in a single device. Conversely, the features and functions of a single device described above can be further divided and embodied by multiple devices.
[0071] The present disclosure further provides a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the method for behavior monitoring according to the first embodiment, the second embodiment, or the third embodiment when running.
[0072] Specifically, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.
[0073] In the description of this specification, reference to the terms "one embodiment," "some embodiments," "example," or "specific example" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification.
[0074] The above description is merely a preferred embodiment of the present disclosure and does not limit the present disclosure in any way. Any technical personnel in the relevant technical field, without departing from the scope of the technical solution of the present disclosure, may make any equivalent substitution, modification, or other changes to the technical solution and technical content disclosed in the present disclosure, and such changes shall be deemed to be within the scope of the technical solution of the present disclosure and shall remain within the scope of protection of the present disclosure.
Claims
1. A method for behavior monitoring, comprising: In the process of monitoring personnel behavior based on image sensors, evaluating personnel behavior based on monitoring data from the image sensors; In response to an assessment result that abnormal behavior exists, obtaining data to be processed associated with the abnormal behavior; Performing privacy processing on the data to be processed; as well as Based on the privacy-processed data and the evaluation results of the abnormal behavior, a personnel behavior event is generated.
2. According to the method of claim 1, obtaining the to-be-processed data associated with the abnormal behavior comprises: Determining the occurrence time of the abnormal behavior, and using the monitoring data starting from the occurrence time of the abnormal behavior as the data to be processed; and / or Determine the evaluation characteristics of the abnormal behavior, and use the monitoring data containing the evaluation characteristics as the data to be processed.
3. According to the method of claim 1, performing privacy processing on the data to be processed comprises: Fuzzy processing is performed on at least a partial area of the data to be processed.
4. According to the method of claim 3, the fuzzy processing of at least a part of the area in the to-be-processed data comprises: The human body region in the data to be processed is subjected to fuzzy processing.
5. According to the method of claim 3, the fuzzy processing of at least a part of the area in the data to be processed comprises: The region where the designated feature in the data to be processed is located is fuzzy processed. The method according to claim 5 , wherein the designated features include features of the person's head and face and / or features of limbs.
7. According to the method of claim 3, the fuzzy processing of at least a part of the area in the data to be processed comprises: Obtaining evaluation characteristics of the abnormal behavior; Determine the location area of the evaluation feature in the data to be processed; as well as Perform fuzzy processing on other areas of the data to be processed except the location area.
8. According to the method of claim 5, the step of performing fuzzy processing on the area where the designated feature in the data to be processed is located comprises: Performing hierarchical blurring processing on the designated features according to the privacy level of the designated features; The privacy level of the specified feature is positively correlated with the level of blurring.
9. The method according to any one of claims 1 to 8, wherein the data to be processed is fuzzy processed by at least one of the following operations: Performing differential processing on the front and back pictures in the data to be processed; Performing edge difference processing on the data to be processed; Reducing the resolution of the data to be processed; Performing size reduction processing on the data to be processed; The data to be processed is subjected to Canny blur processing, Gaussian blur processing and / or average value blur processing.
10. According to the method of any one of claims 1 to 8, generating a personnel behavior event comprises: The privacy-processed data and the evaluation result of the abnormal behavior are fused to obtain the personnel behavior event.
11. The method according to any one of claims 1 to 8, further comprising: Displaying the personnel behavior events; and / or The personnel behavior events are stored locally or in the cloud.
12. An electronic device, characterized in that: include: processor; as well as A memory storing computer instructions for behavior monitoring, wherein when the computer instructions are executed by the processor, the electronic device executes the method according to any one of claims 1-11.
13. A computer-readable storage medium, characterized in that: Contains program instructions for behavior monitoring, and when the program instructions are executed by a processor, the method according to any one of claims 1-11 is implemented.
Citation Information
Patent Citations
Image processing method and device, electronic equipment and storage medium
CN111310605A
Video monitoring method for protecting personnel privacy and system
CN112449155A
Video frame coding method and device, electronic equipment and storage medium
CN113473138A
Double-flow video privacy protection method
CN113965728A
