A same-row relationship detection method and device, electronic equipment and storage medium

Abstract

The application discloses a same-row relationship detection method and device, electronic equipment and a storage medium. The method comprises the following steps: obtaining target video data in a target place; determining a personnel pixel area in each image frame in the target video data, and determining a target height and / or a target body shape ratio of the personnel corresponding to the personnel pixel area; determining an anonymous identity of the personnel in the image frame based on a dynamic anonymous personnel name list corresponding to the target place and the target height and / or the target body shape ratio; and judging whether a first search personnel and a second search personnel are in a same-row relationship according to the anonymous identity information of the personnel in each image frame in the target video data. The scheme can efficiently and with low algorithmic consumption detect the anonymous identity of the personnel in each image frame in the video data, so as to accurately and quickly find out whether the two search personnel are in a same-row relationship according to the anonymous identity of the personnel.

Description

Technical Field

[0001] This invention relates to the field of video surveillance technology, and in particular to a method, apparatus, electronic device, and storage medium for detecting peer relationships. Background Technology

[0002] With the development of video recognition technology, the need to discover potential connections between people through video analysis is becoming increasingly urgent. Discovering peer relationships is a prerequisite for identifying interpersonal relationships and gangs; how to efficiently and with low computational consumption discover peer relationships in videos is a problem that urgently needs to be solved.

[0003] In related technologies, full AI analysis is performed on the video to extract images and feature vectors of individuals. These feature vectors are then compared, and based on the comparison results, individual profiles are generated. These profiles are used to determine the number of times two individuals appear in the same camera within a certain timeframe (e.g., five seconds), and this frequency is used to determine if the two individuals are traveling together. Clearly, this method for determining travel relationships is extremely computationally intensive, and the only evidence of a travel relationship is image data, failing to provide a condensed video record of the individuals traveling together.

[0004] Because camera capture and reporting strategies are typically set to report one image per second (or a period of time), captured images cannot perfectly correspond to the video recordings stored on the central platform at the millisecond level. This discontinuous image evidence leads to a lack of information such as continuous actions and actual duration. Since the captured images reported by the camera cannot completely determine all video frames of the same person in the original video, obtaining condensed evidence of the same person within the original video requires re-analyzing all images in the video using AI, extracting feature vectors, and then using these feature vectors to cluster people and generate personnel profiles. Only in this way can individual or two-person videos of the same person be obtained, which is extremely costly. Summary of the Invention

[0005] This invention provides a method, apparatus, electronic device, and storage medium for detecting peer relationships, which can efficiently and with low computational consumption detect the anonymous identity of people in each image frame of coded video data, thereby accurately and quickly discovering whether two people to be searched have a peer relationship based on the anonymous identity of the people.

[0006] According to one aspect of the present invention, a method for detecting peer relationships is provided, comprising:

[0007] Acquire target video data within the target location;

[0008] For each image frame in the target video data, the pixel region of the person in the image frame is determined based on the target quantization parameters of the macroblock corresponding to each image frame, and the target height and / or target body proportion of the person corresponding to the pixel region of the person is determined.

[0009] The anonymous identity of the person in the image frame is determined based on the dynamic anonymous personnel list database corresponding to the target location and the target's height and / or body proportions.

[0010] Based on the anonymous identity information of the people in each image frame of the target video data, determine whether the first searcher and the second searcher are colleagues.

[0011] According to another aspect of the present invention, a peer relationship detection device is provided, comprising:

[0012] The target video data acquisition module is used to acquire target video data within the target location;

[0013] The personnel information acquisition module is used to determine the personnel pixel region in each image frame of the target video data, and to determine the target height and / or target body proportion of the personnel corresponding to the personnel pixel region.

[0014] The anonymous identity determination module is used to determine the anonymous identity of the person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target's height and / or body proportions.

[0015] The peer relationship determination module is used to determine whether the first searcher and the second searcher are peers based on the anonymous identity information of the people in each image frame of the target video data.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the peer relationship detection method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the peer relationship detection method according to any embodiment of the present invention.

[0021] The peer-to-peer relationship detection scheme of this invention acquires target video data within a target location; for each image frame in the target video data, it determines the pixel region of a person in the image frame and determines the target height and / or target body proportion of the person corresponding to the pixel region; based on the dynamic anonymous personnel list database corresponding to the target location and the target height and / or target body proportion, it determines the anonymous identity of the person in the image frame; and based on the anonymous identity information of the person in each image frame in the target video data, it determines whether the first searcher and the second searcher are peers. Through the technical solution provided by this invention, without performing high-computational operations such as feature vector extraction and feature vector comparison, the anonymous identity of the person in each image frame of the video data can be detected efficiently and with low computational consumption, thereby accurately and quickly discovering whether two searched persons are peers based on their anonymous identities.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 A flowchart of a peer relationship detection method provided in an embodiment of the present invention;

[0025] Figure 2 This is a schematic diagram of camera deployment in a target location provided by an embodiment of the present invention;

[0026] Figure 3 This is a schematic diagram of the structure of a peer relationship detection device provided in an embodiment of the present invention;

[0027] Figure 4 A schematic diagram of the structure of an electronic device for implementing the peer relationship detection method of this invention. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] Figure 1 This is a flowchart illustrating a method for detecting peer relationships according to an embodiment of the present invention. This embodiment is applicable to situations requiring the detection of peer relationships. The method can be executed by a peer relationship detection device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0031] S110. Acquire target video data within the target location.

[0032] The target venue can be an exhibition hall, art gallery, museum, or large shopping mall, among other places. Face / body capture cameras are deployed at the entrances and exits of the target venue. Within the interior areas of the target venue, intelligent coding cameras and face / body capture cameras can be deployed based on the density of people. For example, Figure 2 This is a schematic diagram of camera deployment in a target location provided by an embodiment of the present invention. The intelligent encoding camera is used to collect video data (i.e., recorded data) of the monitored area and encodes the collected video data based on intelligent encoding technology; the face / body capture camera is used to capture images when a target is detected. All intelligent encoding cameras in the target location upload the collected and encoded video data to the central platform, and all face / body capture cameras in the target location also upload the captured image data to the central platform.

[0033] In this embodiment of the invention, the target video data refers to the video data collected and uploaded to the central platform by all intelligent coded cameras within the target location. Optionally, the target video data is data generated by encoding the collected video data using intelligent encoding technology; that is, the target video data is data obtained by encoding the collected video data using intelligent encoding technology. The intelligent encoding technology involves: before encoding, performing target recognition on the image frames in the video data using a deep learning algorithm; during encoding, distinguishing between the target and the background; encoding the target portion using smaller macroblocks and smaller quantization parameters to retain more details; and encoding the background portion using larger macroblocks and larger quantization parameters to save storage space and bandwidth consumption during transmission. The quantization parameter value represents the image quantization step size; the smaller the quantization parameter value, the more details the video image has, and the larger the quantization parameter value, the fewer details the video image has.

[0034] S120. For each image frame in the target video data, determine the person pixel region in the image frame, and determine the target height and / or target body proportion of the person corresponding to the person pixel region.

[0035] In this embodiment of the invention, personnel detection can be performed on each image frame in the target video data to determine the personnel pixel region in the image frame. Optionally, if the target video data is data generated by encoding the collected video data based on intelligent coding technology, then for each image frame in the target video data, the personnel pixel region in the image frame is determined based on the target quantization parameters of the macroblocks corresponding to each image frame. For example, for each image frame in the target video data, the encoding information corresponding to the image frame is used to restore the macroblocks and determine the macroblock map corresponding to the image frame; according to the target quantization parameters of each macroblock in the macroblock map and the quantization parameter features corresponding to the personnel image region, the target macroblock is determined from each macroblock in the macroblock map. Specifically, each macroblock can be traversed based on a preset dimension of the macroblock map. If there are a preset number of consecutive macroblocks whose target quantization parameters are within the quantization parameter range corresponding to the personnel image region, then the preset number of consecutive macroblocks are determined as the target macroblocks; wherein, the preset dimension can be rows or columns. The region formed by connected target macroblocks is taken as the personnel pixel region, or the largest rectangular region contained in the connected target macroblocks is taken as the personnel pixel region. It should be noted that an image frame may include one person, two people, or even multiple people. Therefore, the person pixel region in an image frame may be one or multiple. Based on the shooting time of each image frame in the target video data, the intelligent encoding camera used to capture the image frame, and the person pixel region of the image frame, a list of person pixel regions can be generated. The intelligent encoding camera can be identified by its camera number. For example, Table 1 provides a list of person pixel regions according to an embodiment of the present invention:

[0036] Table 1 List of personnel pixel regions

[0037]

[0038] In this embodiment of the invention, the pixel region of a person is analyzed to determine the target height and / or target body proportion of the person corresponding to the pixel region. For example, based on the installation height and angle of the intelligent encoding camera used to capture the image frame, and the relatively fixed width of the face (distance between cheekbones), the target height and / or body proportion of the person in the pixel region of the image frame is calculated. Optionally, determining the target height and target body proportion of the person corresponding to the pixel region includes: determining the number of pixels occupied by the face width, the number of pixels occupied by the person's height, and the number of pixels occupied by the body width in the pixel region; determining the target height of the person corresponding to the pixel region based on the number of pixels occupied by the face width, the number of pixels occupied by the person's height, and a preset face width; and / or, determining the body width of the person corresponding to the pixel region based on the number of pixels occupied by the face width, the number of pixels occupied by the body width, and the preset face width, and determining the target body proportion of the person corresponding to the pixel region based on the body width and the height.

[0039] For example, assuming a fixed face width, determine the number of pixels occupied by the face width, the number of pixels occupied by the height, and the number of pixels occupied by the body width in the person's pixel region of the image frame. Calculate the target height of the person in the image frame's pixel region using the following formula: Target height = Preset face width / Number of pixels occupied by face width * Number of pixels occupied by height. Calculate the body width of the person in the image frame's pixel region using the following formula: Body width = Preset face width / Number of pixels occupied by face width * Number of pixels occupied by body width. Use the ratio of body width to target height as the target body proportion.

[0040] S130. Determine the anonymous identity of the person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target's height and / or body proportions.

[0041] In this embodiment of the invention, a dynamic anonymous personnel list database corresponding to the target location is obtained. Specifically, since face / body capture cameras are deployed at both the entrance and exit of the target location, images of people entering the location captured by the face / body capture cameras at the entrance can be acquired in real time, and these images can be analyzed to generate a dynamic entry personnel list database. Similarly, images of people leaving the location captured by the face / body capture cameras at the exit can be acquired in real time, and these images can be analyzed to generate a dynamic exit personnel list database. The dynamic entry personnel list database and the dynamic exit personnel list database are compared to generate a dynamic anonymous personnel list database within the target location. The dynamic anonymous personnel list database may include personnel ID, personnel image / feature vector, personnel attributes (whether they wear glasses, whether they wear a hat, clothing color, hair color and length, etc.), height, body proportions, gender, age, entry time, and exit time. For example, Table 2 is a dynamic anonymous personnel list database table provided in this embodiment of the invention:

[0042] Table 2. List of Dynamically Anonymous Individuals

[0043]

[0044] In this embodiment of the invention, a height matching the target height is searched among all heights in the dynamic anonymous personnel list. The personnel ID corresponding to the height matching the target height found in the dynamic anonymous personnel list is used as the anonymous identity of the personnel in the pixel region of the image frame. If all heights in the dynamic anonymous personnel list differ significantly, it indicates that the heights of the personnel in the target location vary considerably. In this case, the personnel ID corresponding to the target height can be quickly found in the dynamic anonymous personnel list. However, if the heights of the personnel in the target location do not differ significantly, that is, if multiple personnel IDs in the dynamic anonymous personnel list match the target height, the personnel ID corresponding to the target body proportion in the dynamic anonymous personnel list can be used as the anonymous identity of the personnel. Optionally, to improve the accuracy of determining the anonymous identity of personnel, the personnel ID corresponding to both the target height and the target body proportion can be searched in the dynamic anonymous personnel list, and the found personnel ID can be used as the anonymous identity of the personnel in the pixel region of the image frame.

[0045] Optionally, before determining the anonymous identity of a person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target height and / or the target body proportion, the method further includes: determining the target capture camera closest to the intelligent encoding camera that captured the image frame; acquiring the target image captured by the target capture camera at the capture time of the image frame; determining the anonymous identity of a person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target height and / or the target body proportion, including: if at least two personnel numbers in the dynamic anonymous personnel list database corresponding to the target location are found to have the same height and body proportion as the target height and the target body proportion, then determining the target personnel number from the at least two personnel numbers; wherein the personnel feature corresponding to the target personnel number matches the personnel feature in the target image; and using the target personnel number as the anonymous identity of the person in the image frame. The advantage of this setting is that the target image captured by the face / body capture camera can help accurately determine the anonymous identity of the person in the pixel region of each image frame in the target video data.

[0046] In this embodiment of the invention, a target capture camera is determined that is closest to the intelligent encoding camera capturing the image frame, and the target image captured by the target capture camera at the capture time of the image frame is obtained. If at least two personnel IDs in the dynamic anonymous personnel list corresponding to the target location are found to have the same height and body proportions as the target height and body proportions, that is, multiple personnel IDs are searched in the dynamic anonymous personnel list based on the target height and body proportions, and it is impossible to specifically determine which personnel ID is the anonymous identity of the personnel pixel area of ​​the image frame, then the personnel ID corresponding to the personnel features that match the personnel features in the target image in the dynamic anonymous personnel list is used as the anonymous identity of the personnel pixel area of ​​the image frame. Here, personnel features may include static attribute information of the face or body, such as hairstyle, whether wearing a mask, gender, age, etc.; personnel features may also include feature images of the face or body. For example, after extracting feature vectors from the target image, the feature vectors extracted from the target image are matched with the feature vectors of each person in the dynamic anonymous personnel list, and the personnel IDs in the dynamic anonymous personnel list whose matching degree exceeds a preset threshold are used as the anonymous identity of the personnel pixel area of ​​the image frame.

[0047] Optionally, determining the anonymous identity of a person in the image frame based on a dynamic anonymous personnel list corresponding to the target location and the target height and / or target body proportion includes: if at least two personnel IDs in the dynamic anonymous personnel list corresponding to the target location are found to have the same height and body proportion as the target height and body proportion, then the predicted position of the personnel corresponding to the at least two personnel IDs at the shooting time of the image frame is determined based on the entry time and / or exit time of the personnel corresponding to the at least two personnel IDs at the target location; from the predicted positions corresponding to the at least two personnel IDs, the target predicted position closest to the location of the intelligent encoding camera that captured the image frame is determined, and the personnel ID corresponding to the target predicted position is used as the anonymous identity of the person in the image frame. The advantage of this setting is that when a personnel ID corresponding to the target height and target body proportion cannot be determined in the dynamic anonymous personnel list, the predicted position at the shooting time of the image frame can be estimated using the entry time and / or exit time of each person in the dynamic anonymous personnel list, thereby assisting in determining the personnel ID of the personnel pixel area in the image frame based on the predicted position and the target height and target body proportion.

[0048] For example, if at least two personnel IDs in the dynamic anonymous personnel list database corresponding to the target location are found to have the same height and body proportions as the target height and body proportions, that is, if multiple personnel IDs are searched in the dynamic anonymous personnel list database based on the target height and body proportions, and it is impossible to specifically determine which personnel ID is the anonymous identity of the personnel in the personnel pixel area of ​​the image frame, then, based on the entry time and / or exit time of each of the at least two personnel IDs at the target location, the predicted position of the personnel corresponding to that personnel ID at the shooting time point of the image frame is determined. From the at least two predicted positions, the target predicted position closest to the location of the intelligent encoding camera that captured the image frame is determined, and the personnel ID corresponding to the target predicted position is used as the anonymous identity of the personnel in the personnel pixel area of ​​the image frame.

[0049] S140. Determine whether the first searcher and the second searcher are colleagues based on the anonymous identity information of the people in each image frame of the target video data.

[0050] In this embodiment of the invention, the anonymous identity information of the person in the pixel region of each image frame in the target video data can be determined through steps S110-S130. Therefore, all image frames containing the anonymous person to be searched can be determined from the target video data, thereby obtaining a condensed video of the anonymous person to be searched. In this embodiment of the invention, it is also possible to determine whether the first searcher and the second searcher are colleagues based on the anonymous identity information of the person in each image frame of the target video data. Optionally, determining whether the first and second searchers are colleagues based on the anonymous identity information of individuals in each image frame of the target video data includes: for each anonymous identity involved in the target video data, determining each target image frame involved in the anonymous identity, and determining the shooting time of each target image frame and the target intelligent encoding camera that shot the target image frame; generating a personnel information table based on the anonymous identity of the personnel, the shooting time of each corresponding target image frame, and the target intelligent encoding camera; searching the personnel information table in chronological order for the first encoding camera that shot the first searcher and the second encoding camera that shot the second searcher at the same time; counting the number of times the first encoding camera and the second encoding camera are the same or adjacent encoding cameras, and determining whether the first and second searchers are colleagues based on the number of times.

[0051] For example, for each anonymous person identity involved in the target video data, each target image frame involved in the anonymous person identity is determined. For example, for the anonymous person identity as Person 1, each target image frame containing Person 1 is determined from the target video data. The shooting time of each target image frame and the target smart encoding camera that shot the target image frame are determined, wherein the target smart encoding camera that shot the target image frame can be characterized by the encoding camera number. Since the installation location of each smart encoding camera in the target location is known, the target smart encoding camera that shot the target image frame is used to reflect the location information of the person corresponding to the anonymous person identity at the shooting time of the target image frame. Based on the anonymous person identity and the shooting time of each corresponding target image frame and the target smart encoding camera, a personnel information table is generated. For example, Table 3 is a personnel information table provided by an embodiment of the present invention:

[0052] Table 3 Personnel Information Table:

[0053] time Personnel Number Encoded camera number Time t1 Personnel 1 Encoded camera n1 Time t1 Personnel 2 Encoded camera n2 Time t1 Personnel 3 Encoded camera n3 ... ... ... ... ... ... time tn Personnel 1 Encoded camera n1 time tn Personnel 2 Encoded camera n2 time tn Personnel 3 Encoded camera n3

[0054] Following the chronological order, in the personnel information table shown in Table 3, locate the first coded camera that captured the first searcher and the second coded camera that captured the second searcher at the same time. For example, if the first searcher is Person 1 and the second searcher is Person 2, then in the personnel information table shown in Table 3, search for the coded camera number corresponding to Person 1 and Person 2 at each time t1, t2, etc. Determine whether the coded camera numbers corresponding to Person 1 and Person 2 at each time are the same or adjacent. If so, it can be determined that Person 1 and Person 2 appeared in the same location at that time. Count the number of times the first and second coded cameras are the same or adjacent, that is, the number of times Person 1 and Person 2 appeared in the same location at the same time. If this number is greater than a preset threshold, it can be determined that Person 1 and Person 2 are in the same line of work; if the number is less than the preset threshold, it can be determined that Person 1 and Person 2 are not in the same line of work.

[0055] The peer-to-peer relationship detection method of this invention acquires target video data within a target location; for each image frame in the target video data, it determines the pixel region of a person in the image frame and determines the target height and / or target body proportion of the person corresponding to the pixel region; based on the dynamic anonymous personnel list database corresponding to the target location and the target height and / or target body proportion, it determines the anonymous identity of the person in the image frame; and based on the anonymous identity information of the person in each image frame in the target video data, it determines whether a first searcher and a second searcher are peers. Through the technical solution provided by this invention, without performing high-computational operations such as feature vector extraction and feature vector comparison, the anonymous identity of the person in each image frame of the video data can be detected efficiently and with low computational consumption, thereby accurately and quickly discovering whether two searched persons are peers based on their anonymous identities.

[0056] Optionally, after determining whether the first and second searchers are in a traveling relationship based on the number of times, the method further includes: if it is determined that the first and second searchers are in a traveling relationship, then extracting the target time when the first and second coded cameras are the same or adjacent coded cameras from the personnel information table; generating a traveling person time camera list based on the target time and the corresponding first and second coded cameras; and generating the traveling trajectory of the first and second searchers based on the traveling person time camera list. The advantage of this setup is that it allows for the rapid determination of the traveling trajectory of two people in a traveling relationship.

[0057] In this embodiment of the invention, if it is determined that the first searcher and the second searcher are colleagues, then the target times where the first coded camera and the second coded camera are the same coded camera or adjacent coded cameras are extracted from the personnel information table. For example, from the personnel information table shown in Table 3, all target times where the first coded camera corresponding to person 1 and the second coded camera corresponding to person 2 are the same coded camera or adjacent coded cameras are determined. A list of colleagues' time cameras is generated based on each target time and the corresponding first and second coded cameras. For example, Table 4 shows a list of colleagues' time cameras provided in this embodiment of the invention:

[0058] Table 4. List of cameras used by fellow travelers

[0059]

[0060] Table 4 shows the list of time cameras for people traveling together, reflecting the locations of people 1 and 2 at various target time points. Therefore, the travel trajectories of people 1 and 2 can be generated based on the list of time cameras for people traveling together.

[0061] Optionally, after generating the travel trajectories of the first and second searchers based on the traveler time camera list, the method further includes: if at least one traveler trajectory point in the traveler trajectory has a trajectory deviation greater than a preset deviation threshold, then filtering for target individuals whose height and body proportions match the traveler trajectory point; obtaining target trajectory points in the target individual's movement trajectory that are at the same time point as the traveler trajectory point, and determining whether the trajectory deviation between the target trajectory point and the traveler trajectory is less than the preset deviation threshold; if so, then using the target individual as the anonymous identity of the person in the image frame corresponding to the traveler trajectory point. For example, in the travel trajectories of individuals 1 and 2, if at a certain time point it is found that the trajectory deviation between the trajectory points of individuals 1 and / or 2 and their travel trajectories is greater than a preset deviation threshold, that is, the trajectory points of individuals 1 and / or 2 at that time point are far from the traveler trajectory, it indicates that there may be an error in determining the anonymous identity of the person in the corresponding image frame at that time point. Therefore, the anonymous identity of the person in the image frame corresponding to the traveler trajectory point is re-determined based on the movement trajectories, height, and body proportions of other individuals. For example, if person 1's trajectory point at time t3 deviates significantly from the peer trajectory, a target person with the same height and body proportions as person 1 is identified from the dynamic anonymous person list database, and the target person's movement trajectory is determined. The target trajectory point at time t3 is determined from the target person's movement trajectory, and it is judged whether the trajectory deviation between the target trajectory point and the peer trajectory is less than a preset deviation threshold. If so, the target person is used as the peer trajectory point to match the anonymous identity of the person in the image frame. For example, if person 3's trajectory point at time t3 is swapped with person 1's trajectory point at time t3, and the trajectories corresponding to person 1 and person 3 are continuous after the swap, then it can be determined that the anonymous identity of person 1 in the image frame corresponding to time t3 should be person 3, and the anonymous identity of person 3 in the image frame corresponding to time t3 should be person 1.

[0062] Optionally, an image frame sequence of a specific person can be extracted from the target video data based on the list of accompanying persons' time-camera recordings, and then played back according to the recording URL. Alternatively, the image frame sequence can be edited to generate a condensed single-person recording URL. Optionally, a condensed video of two accompanying persons can also be extracted from the target video data based on the list of accompanying persons' time-camera recordings. It should be noted that since two accompanying persons may appear not only in the same encoded camera at a certain time point, but also in two adjacent encoded cameras, when two accompanying persons appear in two adjacent encoded cameras, the two image frames captured by the two adjacent encoded cameras at that time point can be played in two separate video playback windows, or the two image frames can be edited and stitched together.

[0063] Figure 3This is a schematic diagram of a parallel relationship detection device provided in an embodiment of the present invention. Figure 3 As shown, the device includes:

[0064] The target video data acquisition module 310 is used to acquire target video data within the target location.

[0065] The personnel information acquisition module 320 is used to determine the personnel pixel region in each image frame of the target video data, and to determine the target height and / or target body proportion of the personnel corresponding to the personnel pixel region.

[0066] Personnel Anonymous Identity Determination Module 330 is used to determine the anonymous identity of personnel in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target height and / or the target body proportions;

[0067] The peer relationship determination module 340 is used to determine whether the first searcher and the second searcher are peers based on the anonymous identity information of the people in each image frame of the target video data.

[0068] Optionally, the peer relationship determination module is used for:

[0069] For each anonymous identity involved in the target video data, determine each target image frame involved in the anonymous identity, and determine the shooting time of each target image frame and the target smart coded camera that shot the target image frame;

[0070] Based on the anonymous identity of the person, the shooting time of each target image frame, and the target intelligent encoding camera, a person information table is generated;

[0071] According to the chronological order, find the first coded camera that took the first searcher's photo and the second coded camera that took the second searcher's photo at the same time in the personnel information table;

[0072] The number of times the first coded camera and the second coded camera are the same or adjacent coded cameras is counted, and the number of times is used to determine whether the first searcher and the second searcher are colleagues.

[0073] Optional, also includes:

[0074] The peer time extraction module is used to extract the target time when the first retrieval personnel and the second retrieval personnel are the same or adjacent retrieval cameras in the personnel information table after determining whether the first retrieval personnel and the second retrieval personnel are peers based on the number of times.

[0075] The time camera list generation module for fellow travelers is used to generate a time camera list for fellow travelers based on the target time and the corresponding first coded camera and second coded camera.

[0076] The peer trajectory generation module is used to generate the peer trajectory of the first searcher and the second searcher based on the peer time camera list.

[0077] Optional, also includes:

[0078] The target personnel screening module is used to filter target personnel whose height and body proportions match those of the target personnel after generating the trajectories of the first and second search personnel based on the time camera list of the trajectories ...

[0079] The anonymous identity determination module obtains the target trajectory point in the movement trajectory of the target person that is at the same time point as the trajectory point of the same person, and determines whether the trajectory deviation between the target trajectory point and the trajectory point of the same person is less than the preset deviation threshold. If so, the target person is identified as the anonymous identity of the person in the image frame corresponding to the trajectory point of the same person.

[0080] Optional, also includes:

[0081] The target capture camera determination module is used to determine the target capture camera that is closest to the intelligent coded camera that captured the image frame before determining the anonymous identity of the person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target height and / or the target body proportion.

[0082] The target image acquisition module is used to acquire the target image captured by the target capture camera at the shooting time point of the image frame;

[0083] The anonymous identity determination module is used for:

[0084] If at least two personnel IDs in the dynamic anonymous personnel list database corresponding to the target location are found to have the same height and body proportions as the target height and body proportions, then the target personnel ID is determined from the at least two personnel IDs; wherein, the personnel characteristics corresponding to the target personnel ID match the personnel characteristics in the target image;

[0085] The target person's ID number is used as the anonymous identity of the person in the image frame.

[0086] Optionally, the personnel anonymity determination module is used for:

[0087] If at least two personnel IDs are found in the dynamic anonymous personnel list database corresponding to the target location, and their height and body proportions are the same as those of the target height and body proportions, then the predicted position of the personnel corresponding to the at least two personnel IDs at the time of the image frame is determined based on their entry time and / or exit time at the target location.

[0088] From the predicted locations corresponding to the at least two personnel IDs, determine the target predicted location that is closest to the location of the smart coded camera that captured the image frame, and use the personnel ID corresponding to the target predicted location as the anonymous identity of the personnel in the image frame.

[0089] Optional, the personnel information acquisition module is used for:

[0090] Determine the number of pixels occupied by the face width, the number of pixels occupied by the height, and the number of pixels occupied by the body width in the personnel pixel region;

[0091] The target height of the person corresponding to the pixel region is determined based on the number of pixels occupied by the face width, the number of pixels occupied by the person's height, and the preset face width; and / or,

[0092] The body width of the person corresponding to the person pixel region is determined based on the number of pixels occupied by the face width, the number of pixels occupied by the body width, and the preset face width. The target body proportion of the person corresponding to the person pixel region is determined based on the body width and the height.

[0093] The peer relationship detection device provided in the embodiments of the present invention can execute the peer relationship detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0094] Figure 4 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0095] like Figure 4As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0096] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0097] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the peer relationship detection method.

[0098] In some embodiments, the peer-to-peer relationship detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the peer-to-peer relationship detection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the peer-to-peer relationship detection method by any other suitable means (e.g., by means of firmware).

[0099] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0100] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0101] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0102] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0103] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0104] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0105] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0106] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for detecting peer relationships, characterized in that, include: Acquire target video data within the target location; For each image frame in the target video data, determine the person pixel region in the image frame, and determine the target height and / or target body proportion of the person corresponding to the person pixel region; The anonymous identity of the person in the image frame is determined based on the dynamic anonymous personnel list database corresponding to the target location and the target's height and / or body proportions. Based on the anonymous identity information of the people in each image frame of the target video data, determine whether the first searcher and the second searcher are colleagues.

2. The method according to claim 1, characterized in that, Determining whether the first and second searchers are colleagues based on the anonymous identity information of people in each image frame of the target video data includes: For each anonymous identity involved in the target video data, determine each target image frame involved in the anonymous identity, and determine the shooting time of each target image frame and the target smart coded camera that shot the target image frame; Based on the anonymous identity of the person, the shooting time of each target image frame, and the target intelligent encoding camera, a person information table is generated; According to the chronological order, find the first coded camera that took the first searcher's photo and the second coded camera that took the second searcher's photo at the same time in the personnel information table; The number of times the first coded camera and the second coded camera are the same or adjacent coded cameras is counted, and the number of times is used to determine whether the first searcher and the second searcher are colleagues.

3. The method according to claim 2, characterized in that, After determining whether the first searcher and the second searcher are in the same industry based on the number of searches, the process further includes: If it is determined that the first searcher and the second searcher are colleagues, then the target time when the first coded camera and the second coded camera are the same or adjacent coded cameras is extracted from the personnel information table. Generate a time camera list of fellow travelers based on the target time and the corresponding first and second coded cameras; Based on the list of time cameras of the people in the same field, the travel trajectories of the first searcher and the second searcher are generated.

4. The method according to claim 3, characterized in that, After generating the travel trajectories of the first and second searchers based on the traveler time camera list, the process further includes: If at least one of the peer trajectory points has a trajectory deviation greater than a preset deviation threshold, then target personnel whose height and body proportions match those of the peer trajectory point will be selected. Obtain the target trajectory point that is at the same time point as the peer trajectory point in the movement trajectory of the target person, and determine whether the trajectory deviation between the target trajectory point and the peer trajectory is less than the preset deviation threshold. If so, the target person is used as the anonymous identity of the person in the image frame corresponding to the peer trajectory point.

5. The method according to claim 1, characterized in that, Before determining the anonymous identity of the person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target's height and / or body proportions, the method further includes: Determine the target capture camera that is closest to the intelligent encoding camera that captured the image frame; Obtain the target image captured by the target capture camera at the capture time point of the image frame; Determining the anonymous identity of individuals in the image frame based on the dynamic anonymous personnel list corresponding to the target location and the target's height and / or body proportions includes: If at least two personnel IDs in the dynamic anonymous personnel list database corresponding to the target location are found to have the same height and body proportions as the target height and body proportions, then the target personnel ID is determined from the at least two personnel IDs; wherein, the personnel characteristics corresponding to the target personnel ID match the personnel characteristics in the target image; The target person's ID number is used as the anonymous identity of the person in the image frame.

6. The method according to claim 1, characterized in that, Determining the anonymous identity of individuals in the image frame based on the dynamic anonymous personnel list corresponding to the target location and the target's height and / or body proportions includes: If at least two personnel IDs are found in the dynamic anonymous personnel list database corresponding to the target location, and their height and body proportions are the same as those of the target height and body proportions, then the predicted position of the personnel corresponding to the at least two personnel IDs at the time of the image frame is determined based on their entry time and / or exit time at the target location. From the predicted locations corresponding to the at least two personnel IDs, determine the target predicted location that is closest to the location of the smart coded camera that captured the image frame, and use the personnel ID corresponding to the target predicted location as the anonymous identity of the personnel in the image frame.

7. The method according to claim 1, characterized in that, Determining the target height and target body proportions of the person corresponding to the pixel region includes: Determine the number of pixels occupied by the face width, the number of pixels occupied by the height, and the number of pixels occupied by the body width in the personnel pixel region; The target height of the person corresponding to the pixel region is determined based on the number of pixels occupied by the face width, the number of pixels occupied by the person's height, and the preset face width; and / or, The body width of the person corresponding to the person pixel region is determined based on the number of pixels occupied by the face width, the number of pixels occupied by the body width, and the preset face width. The target body proportion of the person corresponding to the person pixel region is determined based on the body width and the height.

8. A peer relationship detection device, characterized in that, include: The target video data acquisition module is used to acquire target video data within the target location; The personnel information acquisition module is used to determine the personnel pixel region in each image frame of the target video data, and to determine the target height and / or target body proportion of the personnel corresponding to the personnel pixel region. The anonymous identity determination module is used to determine the anonymous identity of the person in the image frame based on the dynamic anonymous personnel list database corresponding to the target location and the target's height and / or body proportions. The peer relationship determination module is used to determine whether the first searcher and the second searcher are peers based on the anonymous identity information of the people in each image frame of the target video data.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the peer relationship detection method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the peer relationship detection method according to any one of claims 1-7.

Images