A method and device for determining motion trajectory and a computer-readable storage medium

By identifying the close-up images collected by multiple monitoring devices and vehicle features, the problem of missing motion trajectories caused by inconsistent timing in the monitoring system is solved, and a complete pedestrian and vehicle motion trajectories are achieved between multiple devices.

CN112270257BActive Publication Date: 2024-11-08SHENZHEN RENYONG ELECTRONIC SYST CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202011162643.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-10-27
Publication Date
2024-11-08
Estimated Expiration
2040-10-27

AI Technical Summary

Technical Problem

In the existing monitoring system, due to the different timings of the monitoring equipment collecting video images, the movement trajectories of pedestrians and vehicles in the shooting range of multiple devices are missing for a certain period of time, and a complete movement trajectory cannot be obtained.

Method used

By identifying the close-up images collected by multiple monitoring devices, the face feature set and vehicle feature set are obtained, and the motion trajectory of the target person and the vehicle are determined based on these feature sets, and the image timing is consistent. Get the full track.

Benefits of technology

It realizes the acquisition of complete pedestrian and vehicle movement trajectories between multiple monitoring devices, and solves the problem of missing trajectories caused by inconsistent timing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112270257B_ABST
    Figure CN112270257B_ABST
Patent Text Reader

Abstract

The present application is applicable to the field of monitoring and security technology, and provides a method, device and computer-readable storage medium for determining a motion trajectory, wherein the method for determining a motion trajectory comprises: performing facial feature recognition and vehicle feature recognition on N groups of close-up images respectively, and obtaining N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by a monitoring device collecting N different areas at N collection angles within a preset time period; and determining the motion trajectory of a target person and the motion trajectory of a target vehicle based on the comparison results between the N groups of facial feature sets and the comparison results between the N groups of vehicle feature sets. In the above-mentioned method for determining a motion trajectory, since the time sequence of the N groups of close-up images is consistent, the complete motion trajectory of the target person and the target vehicle can be obtained.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of monitoring and security technology, and in particular, relates to a motion trajectory determination method, device and computer-readable storage medium. Background Art

[0002] At present, in order to realize real-time monitoring of some large places, monitoring equipment can be installed in these places. The traditional solution is usually to install a monitoring device at a fixed distance in these large places to realize real-time monitoring of large places, and determine the movement trajectories of pedestrians and vehicles appearing in the multiple monitoring devices through the installed multiple monitoring devices. However, when this solution obtains pedestrian images and / or vehicle images appearing in multiple monitoring devices, due to the different timing of some monitoring devices capturing video images, the movement trajectories of pedestrians and / or vehicles in the shooting range of multiple monitoring devices are missing in a certain period of time, and the complete movement trajectories of pedestrians and / or vehicles cannot be obtained. Summary of the invention

[0003] The embodiments of the present application provide a motion trajectory determination method, device and computer-readable storage medium, which can solve the problem in the existing solution that some of the multiple monitoring devices have different timings for capturing video images, resulting in the motion trajectory of pedestrians and / or vehicles in the shooting range of the multiple monitoring devices being missing in a certain time period, and the complete motion trajectory of the pedestrians and / or vehicles cannot be obtained.

[0004] In a first aspect, an embodiment of the present application provides a method for determining a motion trajectory, comprising:

[0005] Performing facial feature recognition and vehicle feature recognition on N groups of close-up images respectively, to obtain N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by a monitoring device collecting N different areas at N collection angles within a preset time period;

[0006] The motion trajectory of the target person and the motion trajectory of the target vehicle are determined based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets.

[0007] Furthermore, the facial feature recognition and vehicle feature recognition are performed on the N groups of close-up images respectively to obtain N groups of facial feature sets and N groups of vehicle feature sets, including:

[0008] Inputting the N groups of close-up images into a preset first feature extraction model for processing respectively, to obtain N groups of facial feature sets corresponding to the N groups of close-up images;

[0009] The N groups of close-up images are respectively input into a preset second feature extraction model for processing to obtain N groups of vehicle feature sets corresponding to the N groups of close-up images.

[0010] Furthermore, before the N groups of close-up images are respectively input into a preset first feature extraction model for processing, the method further includes:

[0011] Obtain N sets of first image sets;

[0012] Inputting N sets of the first image sets into a preset face detection model and a vehicle detection model respectively, and obtaining a face score value and a vehicle score value corresponding to each first image in the N sets of the first image sets; wherein the face score value is used to indicate the integrity of the face in the first image; and the vehicle score value is used to indicate the integrity of the vehicle in the first image;

[0013] N groups of close-up images are determined according to a preset strategy and a face score value and a vehicle score value corresponding to each first image in the N groups of the first image sets.

[0014] Furthermore, before acquiring the N sets of first image sets, the method further includes:

[0015] Obtain N sets of video image collections;

[0016] Inputting the N groups of video image sets into a preset target detection model for target recognition, and obtaining N groups of target recognition result sets corresponding to the N groups of video image sets;

[0017] The N groups of video image sets including faces and / or vehicles in the N groups of target recognition result sets are determined as the N groups of first image sets.

[0018] Furthermore, the determining the motion trajectory of the target person and the motion trajectory of the target vehicle based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets respectively includes:

[0019] If it is detected that there is a first facial feature set in the N groups of facial feature sets, the movement trajectory of the target person is determined according to the first facial feature set; the first facial feature set is a facial feature set whose similarity with the target facial feature set corresponding to the target person is greater than or equal to a first preset similarity threshold;

[0020] If it is detected that a first vehicle feature set exists in the N groups of vehicle feature sets, the movement trajectory of the target vehicle is determined based on the first vehicle feature set; the first vehicle feature set is a vehicle feature set whose similarity with the target vehicle feature set corresponding to the target vehicle is greater than or equal to a second preset similarity threshold.

[0021] Furthermore, after determining the movement trajectory of the target person and the movement trajectory of the target vehicle, the method further includes:

[0022] The movement trajectory of the target person and the movement trajectory of the target vehicle are sent to a target terminal.

[0023] In a second aspect, an embodiment of the present application provides a motion trajectory determination device, including:

[0024] The recognition unit is used to perform facial feature recognition and vehicle feature recognition on the N groups of close-up images respectively, to obtain N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by collecting N different areas from a monitoring device at N collection angles within a preset time period;

[0025] The first determination unit is used to determine the motion trajectory of the target person and the motion trajectory of the target vehicle based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets.

[0026] In a third aspect, an embodiment of the present application provides a motion trajectory determination device, including:

[0027] A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method for controlling trusted startup as described in any one of the first aspects above when executing the computer program.

[0028] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the motion trajectory determination method as described in any one of the above-mentioned first aspects are implemented.

[0029] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on a motion trajectory determination device, the motion trajectory determination device executes the motion trajectory determination method described in any one of the first aspects above.

[0030] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0031] The embodiment of the present application provides a method for determining a motion trajectory, which obtains N sets of facial feature sets and N sets of vehicle feature sets by performing facial feature recognition and vehicle feature recognition on N sets of close-up images respectively; wherein the N sets of close-up images are obtained by a monitoring device collecting N different areas at N acquisition angles within a preset time period; based on the comparison results between the N sets of facial feature sets and the comparison results between the N sets of vehicle feature sets, the motion trajectory of the target person and the motion trajectory of the target vehicle are determined. In the above-mentioned method for determining the motion trajectory, since the N sets of close-up images are obtained by a monitoring device collecting N different areas at N acquisition angles within a preset time period, the time sequence of the N sets of close-up images is consistent, and the complete motion trajectory of the target person and the target vehicle in N different areas can be obtained by this determination method. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0033] Figure 1 is a flow chart of an implementation method of a motion trajectory determination method provided in an embodiment of the present application;

[0034] Figure 2 is a specific implementation flow chart of S101 in a motion trajectory determination method provided in an embodiment of the present application;

[0035] Figure 3 is a flow chart of an implementation of a motion trajectory determination method provided by another embodiment of the present application;

[0036] Figure 4 is a flow chart of a method for determining a motion trajectory provided in yet another embodiment of the present application;

[0037] Figure 5 is a flow chart of a method for determining a motion trajectory provided by another embodiment of the present application;

[0038] Figure 6 is a structural schematic diagram of a motion trajectory determination device provided in an embodiment of the present application;

[0039] Figure 7 It is a structural schematic diagram of a motion trajectory determination device provided in another embodiment of the present application. DETAILED DESCRIPTION

[0040] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0041] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0042] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0043] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0044] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0045] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0046] See also Figure 1 , Figure 11 is a flow chart of an implementation of a motion trajectory determination method provided in an embodiment of the present application. In an embodiment of the present application, the execution subject of the motion trajectory determination method is a motion trajectory determination device. The motion trajectory determination device may include a terminal or a server, and may also be a chip in a terminal or a processor in a server. Here, the terminal and the server may be a smart phone, a tablet computer, or a desktop computer, etc.

[0047] like Figure 1 As shown, the motion trajectory determination method may include S101 to S102, which are described in detail as follows:

[0048] In S101, facial feature recognition and vehicle feature recognition are performed on N groups of close-up images respectively to obtain N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by a monitoring device collecting N different areas at N collection angles within a preset time period.

[0049] In the embodiment of the present application, the monitoring device may be a billion-pixel camera device, which uses array-type multi-scale ultra-high-definition imaging technology. Array-type multi-scale ultra-high-definition imaging technology is a cross-scale imaging technology with a panoramic main lens + N close-up lens array. Among them, the N close-ups are responsible for clear imaging in their respective areas. The billion-pixel camera device uses the panoramic main lens to achieve positioning and splices the N close-ups together, thereby realizing array-type multi-scale ultra-high-definition imaging.

[0050] Based on this, taking a billion-pixel camera device as an example, N groups of close-up images can be obtained by the billion-pixel camera device capturing N different areas at N acquisition angles within a preset time period. That is to say, the N groups of close-up images correspond to the close-up content in N sub-areas within a main area within a preset time period.

[0051] It should be noted that each group of close-up images is used to describe the close-up content in each sub-area corresponding to each group of close-up images within a preset time period, that is, the time sequence of the N groups of close-up images is consistent, so the motion trajectory determination device can determine the motion trajectory of a pedestrian and / or a vehicle within the preset time period based on the N groups of close-up images. The preset time period can be set according to actual needs and is not limited here.

[0052] In one embodiment of the present application, the motion trajectory determination device may be pre-set with N recognition modules, and the N recognition modules may be controlled to perform facial feature recognition and vehicle feature recognition on N groups of close-up images respectively, thereby improving the working efficiency of the motion trajectory determination device.

[0053] The motion trajectory determination device performs facial feature recognition and vehicle recognition on N groups of close-up images respectively. When N groups of facial feature sets and N groups of vehicle feature sets are obtained, each group of close-up images can be associated with the corresponding facial feature set, and each group of close-up images can be associated with the corresponding vehicle feature set. Among them, the facial feature set and the vehicle feature set can be set according to actual needs, and there is no limitation here. For example, the facial feature set can include but is not limited to the proportion of facial features, whether wearing a mask, whether having a beard, and whether wearing glasses. The vehicle feature set can include but is not limited to the vehicle color, vehicle brand, and license plate number.

[0054] It should be noted that if a close-up image only includes pedestrians, the vehicle feature set of the close-up image is zero; if a close-up image only includes vehicles, the face feature set of the close-up image is zero.

[0055] In the embodiment of the present application, each group of close-up images may include multiple pedestrians and / or multiple vehicles, so the facial feature set corresponding to the group of close-up images may also include multiple facial feature subsets, and the vehicle feature set corresponding to the group of close-up images may also include multiple vehicle feature subsets. In other words, the motion trajectory determination device may associate each pedestrian image in each group of close-up images with the facial feature subset corresponding to the pedestrian image in the facial feature set corresponding to the group of close-up images, and associate each vehicle image in each group of close-up images with the vehicle feature subset corresponding to the vehicle image in the vehicle feature set corresponding to the group of close-up images, so that each facial image in each group of close-up images corresponds to each facial feature subset one-to-one, and each vehicle image in each group of close-up images corresponds to each vehicle feature subset one-to-one.

[0056] In the embodiment of the present application, the motion trajectory determination device pre-stores a plurality of feature extraction models, including: a first feature extraction model and a second feature extraction model.

[0057] The first feature extraction model is used to extract features of pedestrians in the close-up image, and obtain a facial feature set corresponding to each pedestrian in the close-up image. The first feature extraction model can be obtained by training a pre-constructed first deep learning model based on a first preset sample set. Among them, each sample data in the first preset sample set includes a first sample close-up image and a facial feature set corresponding to the first sample close-up image. The value of each element in the facial feature set is used to represent the probability that each pedestrian in the first sample close-up image has a facial feature corresponding to the element. When training the pre-constructed first deep learning model, the first sample close-up image in each sample is used as the input of the first deep learning model, and the facial feature set corresponding to the first sample close-up image in each sample is used as the output of the first deep learning model. Through training, the first deep learning model can learn the correspondence between all possible close-up images and facial feature sets, and the trained first deep learning model is used as the first feature extraction model.

[0058] For example, assuming that the facial feature set is [A1, B1, C1], element A1, element B1 and element C1 correspond to facial feature 1, facial feature 2 and facial feature 3 respectively, then the value of element A1 is used to represent the probability of each pedestrian in each group of close-up images having facial feature 1, the value of element B1 is used to represent the probability of each pedestrian in each group of close-up images having facial feature 2, and the value of element C1 is used to represent the probability of each pedestrian in each group of close-up images having facial feature 3.

[0059] The second feature extraction model is used to extract features of pedestrians in the close-up image, and obtain a vehicle feature set corresponding to each vehicle in the close-up image. The second feature extraction model can be obtained by training a pre-constructed second deep learning model based on a second preset sample set. Among them, each sample data in the second preset sample set includes a second sample close-up image and a vehicle feature set corresponding to the second sample close-up image. The value of each element in the vehicle feature set is used to represent the probability that each vehicle in the second sample close-up image has a vehicle feature corresponding to the element. When training the pre-constructed second deep learning model, the second sample close-up image in each sample is used as the input of the second deep learning model, and the vehicle feature set corresponding to the second sample close-up image in each sample is used as the output of the second deep learning model. Through training, the second deep learning model can learn the correspondence between all possible close-up images and vehicle feature sets, and the trained second deep learning model is used as the second feature extraction model.

[0060] For example, assuming that the vehicle feature set is [A2, B2, C2], element A2, element B2 and element C2 correspond to vehicle feature 1, vehicle feature 2 and vehicle feature 3 respectively, then the value of element A2 is used to represent the probability of the existence of vehicle feature 1 for each vehicle in each group of close-up images, the value of element B2 is used to represent the probability of the existence of vehicle feature 2 for each vehicle in each group of close-up images, and the value of element C2 is used to represent the probability of the existence of vehicle feature 3 for each vehicle in each group of close-up images.

[0061] Based on this, in one embodiment of the present application, the motion trajectory determination device can specifically obtain N sets of face sets and N sets of vehicle feature sets through the following steps S201 to S202, as described in detail as follows:

[0062] In S201, the N groups of close-up images are respectively input into a preset first feature extraction model for processing to obtain N groups of facial feature sets corresponding to the N groups of close-up images.

[0063] In this embodiment, the motion trajectory determination device may also pre-store a plurality of detection models, including: a face detection model and a vehicle detection model.

[0064] The face detection model is used to detect the integrity of the face contained in the first image. The face detection model can be obtained by training a pre-constructed third deep learning model based on a third preset sample set. Among them, each sample data in the third preset sample set includes a first sample image and a face score value corresponding to the first sample image. The face score value is used to represent the completeness and clarity of the facial features in the face of each pedestrian in the first sample image. The higher the score value, the more complete the facial features are and the higher the clarity of the facial features. When training the pre-constructed third deep learning model, the first sample image in each sample is used as the input of the third deep learning model, and the face score value corresponding to the first sample image in each sample is used as the output of the third deep learning model. Through training, the third deep learning model can learn the correspondence between all possible sample images and face score values, and the trained third deep learning model is used as the face detection model.

[0065] The vehicle detection model is used to detect the integrity of the vehicle contained in the first image. The vehicle detection model can be obtained by training the pre-constructed fourth deep learning model based on the fourth preset sample set. Among them, each sample data in the fourth preset sample set includes a second sample image and a vehicle score value corresponding to the second sample image. The vehicle score value is used to represent the body integrity and clarity of each vehicle in the second sample image. The higher the score value, the more complete the body and the higher the clarity of the body. When training the pre-constructed fourth deep learning model, the second sample image in each sample is used as the input of the fourth deep learning model, and the vehicle score value corresponding to the second sample image in each sample is used as the output of the fourth deep learning model. Through training, the fourth deep learning model can learn the correspondence between all possible sample images and vehicle score values, and the trained fourth deep learning model is used as the vehicle detection model.

[0066] Based on this, in another embodiment of the present application, the motion trajectory determination device can be Figure 3 S301 to S303 shown determine N groups of close-up images, as detailed below:

[0067] In S301, N groups of first image sets are obtained.

[0068] In this embodiment, since the video image may include other living things and / or inanimate objects in addition to pedestrians and vehicles, the motion trajectory determination device may also preset and store a target detection model. The target detection model is used to detect the target object in the image and identify the type of the target object. The target detection model may be an existing target detection model based on a convolutional neural network.

[0069] Based on this, in another embodiment of the present application, the motion trajectory determination device can be as follows: Figure 4 S401 to S403 shown determine that the video image set is the first image set, which is described in detail as follows:

[0070] In S401, N sets of video images are obtained.

[0071] In S402, the N groups of video image sets are input into a preset target detection model for target recognition, and N groups of target recognition result sets corresponding to the N groups of video image sets are obtained.

[0072] In S403, the N groups of video image sets including faces and / or vehicles in the N groups of target recognition result sets are determined as N groups of first image sets.

[0073] In this embodiment, the target recognition result set is used to describe the type of the target object contained in each video image in the video image set.

[0074] The types of target objects may include but are not limited to pedestrians, vehicles and other non-pedestrians. For example, non-pedestrians may include but are not limited to cats, dogs, mice or buildings.

[0075] When the target recognition result corresponding to a certain video image includes pedestrians and / or vehicles, it means that the video image includes pedestrians and / or vehicles. Therefore, the motion trajectory determination device can determine the video image set including pedestrians and / or vehicles in the target recognition result set as the first image set.

[0076] In S302, N groups of the first image sets are respectively input into a preset face detection model and a vehicle detection model to obtain a face score value and a vehicle score value corresponding to each first image in the N groups of the first image sets; wherein the face score value is used to indicate the integrity of the face in the first image; and the vehicle score value is used to indicate the integrity of the vehicle in the first image.

[0077] In S303, N groups of close-up images are determined according to a preset strategy and a face score value and a vehicle score value corresponding to each first image in the N groups of the first image sets.

[0078] In this embodiment, the preset strategy can be set according to actual needs and is not limited here. Exemplarily, the preset strategy may include three sub-strategies: the first sub-strategy may be: if the first image set only includes pedestrians, then the close-up image of each group is determined according to the face score values ​​corresponding to each of the first image set; the second sub-strategy may be: if the first image set only includes vehicles, then the close-up image of each group is determined according to the vehicle score values ​​corresponding to each of the first image set; the third sub-strategy may be: if the first image set includes pedestrians and vehicles, then the close-up image of each group is determined according to the face score values ​​and vehicle score values ​​corresponding to each of the first image set.

[0079] Specifically, if the motion trajectory determination device detects that the first image set only includes pedestrians, the first image with the largest face score value is determined as the close-up image of each group; if the motion trajectory determination device detects that the first image set only includes vehicles, the first image with the largest vehicle score value is determined as the close-up image of each group; if the motion trajectory determination device detects that the first image set includes pedestrians and vehicles, the first image with a face score value greater than a face score value threshold and a vehicle score value greater than a vehicle score value threshold is determined as the close-up image of each group.

[0080] After the motion trajectory determination device determines N groups of close-up images, the N groups of close-up images can be respectively input into a pre-constructed first feature extraction model for processing to obtain N groups of facial feature sets corresponding to the N groups of close-up images.

[0081] In this embodiment, after obtaining N sets of facial feature sets corresponding to N sets of close-up images, the motion trajectory determination device may compare the value of each element in each set of facial feature sets with the first preset probability threshold corresponding to the element. If the motion trajectory determination device detects that the value of at least one element in the facial feature set is greater than the first preset probability threshold corresponding to the element, it is determined that the pedestrian in the close-up image corresponding to the set of facial feature sets has the facial feature corresponding to the element.

[0082] Exemplarily, assuming that the value of the facial feature set [A1, B1, C1] is [0.9, 0.7, 0.95], and the first preset probability threshold corresponding to each element is 0.9, since the probability value of element C1 is greater than 0.9, the motion trajectory determination device can determine that the pedestrian in the close-up image corresponding to the group of facial feature sets has facial feature 3 corresponding to element C1.

[0083] It should be noted that the first preset probability threshold corresponding to each element may be the same or different, and the first preset probability threshold may be set according to actual needs and is not limited here.

[0084] The motion trajectory determination device inputs each group of close-up images into a pre-constructed first feature extraction model for processing, which may specifically include: inputting each group of close-up images into a plurality of first feature extraction sub-models for detecting different facial features, each first feature extraction sub-model outputs a probability value, and then obtains a facial feature set composed of each probability value. Each probability value is used to represent the probability of a certain facial feature existing in the pedestrian in the close-up image.

[0085] In S202, the N groups of close-up images are respectively input into a preset second feature extraction model for processing to obtain N groups of vehicle feature sets corresponding to the N groups of close-up images.

[0086] In this embodiment, after obtaining N sets of vehicle feature sets corresponding to N sets of close-up images, the motion trajectory determination device may compare the value of each element in each set of vehicle feature sets with the second preset probability threshold corresponding to the element. If the motion trajectory determination device detects that the value of at least one element in the vehicle feature set is greater than the second preset probability threshold corresponding to the element, it is determined that the vehicle in the close-up image corresponding to the set of vehicle feature sets has the vehicle feature corresponding to the element.

[0087] Exemplarily, assuming that the value of the vehicle feature set [A2, B2, C2] is [0.9, 0.7, 0.95], and the second preset probability threshold corresponding to each element is 0.9, since the probability value of element C1 is greater than 0.9, the motion trajectory determination device can determine that the vehicle in the close-up image corresponding to the group of vehicle feature sets has vehicle feature 3 corresponding to element C2.

[0088] It should be noted that the second preset probability threshold corresponding to each element may be the same or different, and the second preset probability threshold may be set according to actual needs and is not limited here.

[0089] The motion trajectory determination device inputs each group of close-up images into a pre-constructed second feature extraction model for processing, which may specifically include: inputting each group of close-up images into a plurality of second feature extraction sub-models for detecting different vehicle features, each second feature extraction sub-model outputs a probability value, and then obtains a vehicle feature set composed of each probability value. Each probability value is used to represent the probability of a certain vehicle feature existing in the vehicle in the close-up image.

[0090] In S102, the motion trajectory of the target person and the motion trajectory of the target vehicle are determined based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets.

[0091] In an embodiment of the present application, after obtaining N groups of facial feature sets and N groups of vehicle feature sets, the motion trajectory determination device can compare each group of facial feature sets one by one, and compare each group of vehicle feature sets one by one, that is, compare each facial feature subset in each group of facial feature sets with each facial feature subset in other groups of facial feature sets one by one, and compare each vehicle feature subset in each vehicle feature set with each vehicle feature subset in other groups of vehicle feature sets one by one, to obtain the comparison results of each facial feature subset in each group of facial feature sets with each facial feature subset in other groups of facial feature sets and the comparison results of each vehicle feature subset in each group of vehicle feature sets with each vehicle feature subset in other groups of vehicle feature sets, and determine the motion trajectories of the target person and the target vehicle based on the above comparison results.

[0092] Specifically, in another embodiment of the present application, the motion trajectory determination device can be Figure 5 Steps S501 to S502 shown determine the movement trajectories of the target person and the target vehicle, as detailed below:

[0093] In S501, if it is detected that there is a first facial feature set in the N groups of facial feature sets, the movement trajectory of the target person is determined based on the first facial feature set; the first facial feature set is a facial feature set whose similarity with the target facial feature set corresponding to the target person is greater than or equal to a first preset similarity threshold.

[0094] In this embodiment, combined with S102, each group of facial feature sets includes multiple facial feature subsets. Therefore, the motion trajectory determination device can determine the pedestrian image corresponding to any facial feature subset in any group of facial feature sets as the target person, and determine the facial feature subset as the target facial feature set of the target person.

[0095] Based on this, the motion trajectory determination device can compare the facial feature subsets in the remaining N-1 facial feature sets except the facial feature set with the target facial feature set corresponding to the target person one by one. If the motion trajectory determination device detects that there is a first facial feature set corresponding to the first facial feature subset whose similarity with the target facial feature set is greater than or equal to the first preset similarity threshold among the facial feature subsets in the remaining N-1 facial feature sets, the pedestrian in the pedestrian image corresponding to the first facial feature set is determined as the target person, indicating that the target person has reached the first sub-area corresponding to the close-up image corresponding to the first facial feature set from the target sub-area corresponding to the close-up image corresponding to the target facial feature set, thereby determining the motion trajectory of the target person. Among them, the first preset similarity threshold can be set according to actual needs and is not limited here.

[0096] It should be noted that, in this embodiment, the similarity between the target facial feature set corresponding to the target person and the first facial feature set is greater than or equal to the first preset similarity threshold, specifically referring to the similarity between each facial feature included in the target facial feature set and each corresponding facial feature included in the first facial feature set is greater than or equal to the first preset similarity threshold.

[0097] In S502, if it is detected that there is a first vehicle feature set in the N groups of vehicle feature sets, the movement trajectory of the target vehicle is determined based on the first vehicle feature set; the first vehicle feature set is a vehicle feature set whose similarity with the target vehicle feature set corresponding to the target vehicle is greater than or equal to a second preset similarity threshold.

[0098] In this embodiment, in combination with S102, each group of vehicle feature sets includes multiple vehicle feature subsets. Therefore, the motion trajectory determination device can determine the pedestrian image corresponding to any vehicle feature subset in any group of vehicle feature sets as the target vehicle, and determine the vehicle feature subset as the target vehicle feature set of the target vehicle.

[0099] Based on this, the motion trajectory determination device can compare the vehicle feature subsets in the remaining N-1 vehicle feature sets except the vehicle feature set with the target vehicle feature set corresponding to the target vehicle one by one. If the motion trajectory determination device detects that there is a first vehicle feature set corresponding to the first vehicle feature subset in the remaining N-1 vehicle feature sets whose similarity with the target vehicle feature set is greater than or equal to the second preset similarity threshold, the pedestrian in the pedestrian image corresponding to the first vehicle feature set is determined as the target vehicle, indicating that the target vehicle has reached the first sub-area corresponding to the close-up image corresponding to the first vehicle feature set from the target sub-area corresponding to the close-up image corresponding to the target vehicle feature set, thereby determining the motion trajectory of the target vehicle. Among them, the second preset similarity threshold can be set according to actual needs and is not limited here.

[0100] It should be noted that, in the present embodiment, the similarity between the target vehicle feature set corresponding to the target vehicle and the first vehicle feature set is greater than or equal to the second preset similarity threshold, specifically referring to that the similarity between each vehicle feature included in the target vehicle feature set and each corresponding vehicle feature included in the first vehicle feature set is greater than or equal to the second preset similarity threshold.

[0101] In another embodiment of the present application, after determining the movement trajectory of the target person and the movement trajectory of the target vehicle, the movement trajectory of the target person and the movement trajectory of the target vehicle can also be sent to the target terminal, and the target terminal stores the movement trajectory of the target person and the movement trajectory of the target vehicle, so that relevant personnel can determine the movement trajectory of the target person and the target vehicle through the target terminal. The target terminal can be a tablet computer or a desktop computer, etc.

[0102] In another embodiment of the present application, the motion trajectory determination device can also associate the motion trajectory of the target person with the target facial feature set of the target person, and send it to the target terminal; associate the motion trajectory of the target vehicle with the target vehicle feature set of the target vehicle, and send it to the target terminal, so that relevant personnel can browse the motion trajectory of any pedestrian or vehicle through the facial feature set and the vehicle feature set.

[0103] From the above, it can be seen that the present embodiment provides a method for determining a motion trajectory, which obtains N sets of facial feature sets and N sets of vehicle feature sets by performing facial feature recognition and vehicle feature recognition on N sets of close-up images respectively; wherein the N sets of close-up images are obtained by a monitoring device collecting N different areas at N acquisition angles within a preset time period; and the motion trajectory of the target person and the motion trajectory of the target vehicle are determined based on the comparison results between the N sets of facial feature sets and the comparison results between the N sets of vehicle feature sets. In the above method for determining the motion trajectory, since the N sets of close-up images are obtained by a monitoring device collecting N different areas at N acquisition angles within a preset time period, the time sequence of the N sets of close-up images is consistent, and the determination method can obtain the complete motion trajectory of the target person and the target vehicle in N different areas.

[0104] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0105] Corresponding to the motion trajectory determination method described in the above embodiment, Figure 6 The structure block diagram of a motion trajectory determination device provided by an embodiment of the present application is shown. For the convenience of explanation, only the part related to the embodiment of the present application is shown. Figure 6 The motion trajectory determination device 600 includes: an identification unit 61 and a first determination unit 62. Wherein:

[0106] The recognition unit 61 is used to perform facial feature recognition and vehicle feature recognition on N groups of close-up images respectively, to obtain N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by a monitoring device collecting N different areas at N collection angles within a preset time period.

[0107] The first determination unit 62 is used to determine the motion trajectory of the target person and the motion trajectory of the target vehicle based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets.

[0108] In one embodiment of the present application, the identification unit 62 specifically includes: a first processing unit and a second processing unit.

[0109] The first processing unit is used to input the N groups of close-up images into a preset first feature extraction model for processing, so as to obtain N groups of facial feature sets corresponding to the N groups of close-up images.

[0110] The second processing unit is used to input the N groups of close-up images into a preset second feature extraction model for processing, so as to obtain N groups of vehicle feature sets corresponding to the N groups of close-up images.

[0111] In one embodiment of the present application, the motion trajectory determination device 600 further includes: a first acquisition unit, a first detection unit and a second determination unit.

[0112] The first acquisition unit is used to acquire N groups of first image sets.

[0113] The first detection unit is used to input N groups of the first image sets into a preset face detection model and a vehicle detection model respectively, to obtain a face score value and a vehicle score value corresponding to each first image in the N groups of the first image sets; wherein the face score value is used to indicate the integrity of the face in the first image; and the vehicle score value is used to indicate the integrity of the vehicle in the first image.

[0114] The second determination unit is used to determine N groups of close-up images according to a preset strategy and a face score value and a vehicle score value corresponding to each first image in the N groups of first image sets.

[0115] In one embodiment of the present application, the motion trajectory determination device 600 further includes: a second acquisition unit, a target recognition unit and a third determination unit.

[0116] The second acquisition unit is used to acquire N sets of video image sets.

[0117] The target recognition unit is used to input the N groups of video image sets into a preset target detection model for target recognition, and obtain N groups of target recognition result sets corresponding to each of the N groups of video image sets.

[0118] The third determination unit is used to determine the N groups of video image sets including faces and / or vehicles in the N groups of target recognition result sets as the N groups of first image sets.

[0119] In one embodiment of the present application, the first determining unit 62 further includes: a fourth determining unit and a fifth determining unit. Among them:

[0120] The fourth determination unit is used to determine the movement trajectory of the target person based on the first facial feature set if it is detected that there is a first facial feature set in the N groups of facial feature sets; the first facial feature set is a facial feature set whose similarity with the target facial feature set corresponding to the target person is greater than or equal to a first preset similarity threshold.

[0121] The fifth determination unit is used to determine the movement trajectory of the target vehicle according to the first vehicle feature set if it is detected that there is a first vehicle feature set in the N groups of vehicle feature sets; the first vehicle feature set is a vehicle feature set whose similarity with the target vehicle feature set corresponding to the target vehicle is greater than or equal to a second preset similarity threshold.

[0122] In one embodiment of the present application, the motion trajectory determination device further includes: a sending unit.

[0123] The sending unit is used to send the motion trajectory of the target person and the motion trajectory of the target vehicle to the target terminal.

[0124] From the above, it can be seen that the motion trajectory determination device provided by the embodiment of the present application obtains N sets of facial feature sets and N sets of vehicle feature sets by performing facial feature recognition and vehicle feature recognition on N sets of close-up images respectively; wherein the N sets of close-up images are obtained by a monitoring device collecting N different areas at N acquisition angles within a preset time period; based on the comparison results between the N sets of facial feature sets and the comparison results between the N sets of vehicle feature sets, the motion trajectory of the target person and the motion trajectory of the target vehicle are determined respectively. In the above motion trajectory determination method, since the N sets of close-up images are obtained by a monitoring device collecting N different areas at N acquisition angles within a preset time period, the time sequence of the N sets of close-up images is consistent, and the complete motion trajectory of the target person and the target vehicle in N different areas can be obtained by this determination method.

[0125] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0126] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0127] Figure 7 This is a schematic diagram of the structure of a motion trajectory determination device provided in one embodiment of the present application. Figure 7 As shown, the motion trajectory determination device 7 of this embodiment includes: at least one processor 70 ( Figure 7 Only one is shown in the figure) a processor, a memory 71, and a computer program 72 stored in the memory 71 and executable on the at least one processor 70, and when the processor 70 executes the computer program 72, the steps in any of the above-mentioned embodiments of the method for determining the motion trajectory are implemented.

[0128] The motion trajectory determination device 7 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The motion trajectory determination device may include, but is not limited to, a processor 70 and a memory 71. Those skilled in the art will appreciate that Figure 7 It is only an example of the motion trajectory determination device 7 and does not constitute a limitation on the motion trajectory determination device 7. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, etc.

[0129] The processor 70 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0130] In some embodiments, the memory 71 may be an internal storage unit of the motion trajectory determination device 7, such as a hard disk or memory of the motion trajectory determination device 7. In other embodiments, the memory 71 may also be an external storage device of the motion trajectory determination device 7, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the motion trajectory determination device 7. Further, the memory 71 may also include both an internal storage unit and an external storage device of the motion trajectory determination device 7. The memory 71 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory 71 may also be used to temporarily store data that has been output or is to be output.

[0131] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0132] An embodiment of the present application provides a computer program product. When the computer program product runs on a motion trajectory determination device, the motion trajectory determination device can implement the steps in the above-mentioned various method embodiments when executed.

[0133] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device capable of carrying the computer program code to the terminal device, a recording medium, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), an electric carrier signal, a telecommunication signal and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0134] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0135] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0136] In the embodiments provided in the present application, it should be understood that the disclosed motion trajectory determination device and method can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0137] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0138] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for determining a motion trajectory, characterized in that: include: The motion trajectory determination device performs facial feature recognition and vehicle feature recognition on N groups of close-up images respectively, and obtains N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by a monitoring device collecting N different areas at N collection angles within a preset time period; Determining a motion trajectory of a target person and a motion trajectory of a target vehicle based on comparison results between the N groups of face feature sets and comparison results between the N groups of vehicle feature sets respectively; The motion trajectory determination device pre-stores a plurality of feature extraction models, including: a first feature extraction model and a second feature extraction model; the first feature extraction model is used to extract features of pedestrians in a close-up image to obtain a facial feature set corresponding to each pedestrian in the close-up image; the first feature extraction model is obtained by training a pre-constructed first deep learning model based on a first preset sample set; each piece of sample data in the first preset sample set includes a first sample close-up image and a facial feature set corresponding to the first sample close-up image; the value of each element in the facial feature set is used to represent the probability that each pedestrian in the first sample close-up image has a facial feature corresponding to the element; The second feature extraction model is used to extract features of vehicles in the close-up image to obtain a vehicle feature set corresponding to each vehicle in the close-up image; the second feature extraction model is obtained by training a pre-constructed second deep learning model based on a second preset sample set; wherein each sample data in the second preset sample set includes a second sample close-up image and a vehicle feature set corresponding to the second sample close-up image; the value of each element in the vehicle feature set is used to represent the probability that each vehicle in the second sample close-up image has a vehicle feature corresponding to the element; The facial feature recognition and vehicle feature recognition are performed on the N groups of close-up images respectively to obtain N groups of facial feature sets and N groups of vehicle feature sets, including: Inputting the N groups of close-up images into a preset first feature extraction model for processing respectively, to obtain N groups of facial feature sets corresponding to the N groups of close-up images; Inputting the N groups of close-up images into a preset second feature extraction model for processing respectively, to obtain N groups of vehicle feature sets corresponding to the N groups of close-up images; The motion trajectory determination device also pre-stores multiple detection models, including: a face detection model and a vehicle detection model; the face detection model is used to detect the integrity of the face contained in the first image; the face detection model is obtained by training a pre-constructed third deep learning model based on a third preset sample set; wherein each sample data in the third preset sample set includes a first sample image and a face score value corresponding to the first sample image; the face score value is used to indicate the completeness and clarity of the facial features in the face of each pedestrian in the first sample image, and the higher the score value, the more complete the facial features are, and the higher the clarity of the facial features is; the vehicle detection model is used to detect the integrity of the vehicle contained in the first image; the vehicle detection model is obtained by training a pre-constructed fourth deep learning model based on a fourth preset sample set; wherein each sample data in the fourth preset sample set includes a second sample image and a vehicle score value corresponding to the second sample image; the vehicle score value is used to indicate the body integrity and clarity of each vehicle in the second sample image, and the higher the score value, the more complete the body is, and the higher the clarity of the body is; Before the N groups of close-up images are respectively input into a preset first feature extraction model for processing, the method further includes: Obtain N sets of first image sets; Inputting N sets of the first image sets into a preset face detection model and a vehicle detection model respectively, and obtaining a face score value and a vehicle score value corresponding to each first image in the N sets of the first image sets; wherein the face score value is used to indicate the integrity of the face in the first image; and the vehicle score value is used to indicate the integrity of the vehicle in the first image; Determining N groups of close-up images according to a preset strategy and a face score value and a vehicle score value corresponding to each first image in the N groups of first image sets includes: If the motion trajectory determination device detects that the first image set includes only pedestrians, the first image with the largest face score value is determined as the close-up image of each group; if the motion trajectory determination device detects that the first image set includes only vehicles, the first image with the largest vehicle score value is determined as the close-up image of each group; if the motion trajectory determination device detects that the first image set includes pedestrians and vehicles, the first image with a face score value greater than a face score value threshold and a vehicle score value greater than a vehicle score value threshold is determined as the close-up image of each group; The determining the motion trajectory of the target person and the motion trajectory of the target vehicle based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets respectively includes: If it is detected that there is a first facial feature set in the N groups of facial feature sets, the movement trajectory of the target person is determined according to the first facial feature set; the first facial feature set is a facial feature set whose similarity with the target facial feature set corresponding to the target person is greater than or equal to a first preset similarity threshold; If it is detected that a first vehicle feature set exists in the N groups of vehicle feature sets, the movement trajectory of the target vehicle is determined based on the first vehicle feature set; the first vehicle feature set is a vehicle feature set whose similarity with the target vehicle feature set corresponding to the target vehicle is greater than or equal to a second preset similarity threshold.

2. The motion trajectory determination method according to claim 1, characterized in that: Before acquiring the N sets of first image sets, the method further includes: Obtain N sets of video image collections; Inputting the N groups of video image sets into a preset target detection model for target recognition, and obtaining N groups of target recognition result sets corresponding to the N groups of video image sets; The N groups of video image sets including faces and / or vehicles in the N groups of target recognition result sets are determined as the N groups of first image sets.

3. The motion trajectory determination method according to claim 1, characterized in that: After determining the movement trajectory of the target person and the movement trajectory of the target vehicle, the method further includes: The movement trajectory of the target person and the movement trajectory of the target vehicle are sent to a target terminal.

4. A motion trajectory determination device, characterized in that: include: The recognition unit is used for the motion trajectory determination device to perform facial feature recognition and vehicle feature recognition on N groups of close-up images respectively, to obtain N groups of facial feature sets and N groups of vehicle feature sets; wherein the N groups of close-up images are obtained by a monitoring device collecting N different areas at N collection angles within a preset time period; A first determining unit is used to determine a motion trajectory of a target person and a motion trajectory of a target vehicle based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets respectively; The motion trajectory determination device pre-stores a plurality of feature extraction models, including: a first feature extraction model and a second feature extraction model; the first feature extraction model is used to extract features of pedestrians in a close-up image to obtain a facial feature set corresponding to each pedestrian in the close-up image; the first feature extraction model is obtained by training a pre-constructed first deep learning model based on a first preset sample set; each piece of sample data in the first preset sample set includes a first sample close-up image and a facial feature set corresponding to the first sample close-up image; the value of each element in the facial feature set is used to represent the probability that each pedestrian in the first sample close-up image has a facial feature corresponding to the element; The second feature extraction model is used to extract features of vehicles in the close-up image to obtain a vehicle feature set corresponding to each vehicle in the close-up image; the second feature extraction model is obtained by training a pre-constructed second deep learning model based on a second preset sample set; wherein each sample data in the second preset sample set includes a second sample close-up image and a vehicle feature set corresponding to the second sample close-up image; the value of each element in the vehicle feature set is used to represent the probability that each vehicle in the second sample close-up image has a vehicle feature corresponding to the element; The facial feature recognition and vehicle feature recognition are performed on the N groups of close-up images respectively to obtain N groups of facial feature sets and N groups of vehicle feature sets, including: Inputting the N groups of close-up images into a preset first feature extraction model for processing respectively, to obtain N groups of facial feature sets corresponding to the N groups of close-up images; Inputting the N groups of close-up images into a preset second feature extraction model for processing respectively, to obtain N groups of vehicle feature sets corresponding to the N groups of close-up images; The motion trajectory determination device also pre-stores multiple detection models, including: a face detection model and a vehicle detection model; the face detection model is used to detect the integrity of the face contained in the first image; the face detection model is obtained by training a pre-constructed third deep learning model based on a third preset sample set; wherein each sample data in the third preset sample set includes a first sample image and a face score value corresponding to the first sample image; the face score value is used to indicate the completeness and clarity of the facial features in the face of each pedestrian in the first sample image, and the higher the score value, the more complete the facial features are, and the higher the clarity of the facial features is; the vehicle detection model is used to detect the integrity of the vehicle contained in the first image; the vehicle detection model is obtained by training a pre-constructed fourth deep learning model based on a fourth preset sample set; wherein each sample data in the fourth preset sample set includes a second sample image and a vehicle score value corresponding to the second sample image; the vehicle score value is used to indicate the body integrity and clarity of each vehicle in the second sample image, and the higher the score value, the more complete the body is, and the higher the clarity of the body is; Before the N groups of close-up images are respectively input into a preset first feature extraction model for processing, the method further includes: Obtain N sets of first image sets; Inputting N sets of the first image sets into a preset face detection model and a vehicle detection model respectively, and obtaining a face score value and a vehicle score value corresponding to each first image in the N sets of the first image sets; wherein the face score value is used to indicate the integrity of the face in the first image; and the vehicle score value is used to indicate the integrity of the vehicle in the first image; Determining N groups of close-up images according to a preset strategy and a face score value and a vehicle score value corresponding to each first image in the N groups of first image sets includes: If the motion trajectory determination device detects that the first image set includes only pedestrians, the first image with the largest face score value is determined as the close-up image of each group; if the motion trajectory determination device detects that the first image set includes only vehicles, the first image with the largest vehicle score value is determined as the close-up image of each group; if the motion trajectory determination device detects that the first image set includes pedestrians and vehicles, the first image with a face score value greater than a face score value threshold and a vehicle score value greater than a vehicle score value threshold is determined as the close-up image of each group; The determining the motion trajectory of the target person and the motion trajectory of the target vehicle based on the comparison results between the N groups of face feature sets and the comparison results between the N groups of vehicle feature sets respectively includes: If it is detected that there is a first facial feature set in the N groups of facial feature sets, the movement trajectory of the target person is determined according to the first facial feature set; the first facial feature set is a facial feature set whose similarity with the target facial feature set corresponding to the target person is greater than or equal to a first preset similarity threshold; If it is detected that a first vehicle feature set exists in the N groups of vehicle feature sets, the movement trajectory of the target vehicle is determined based on the first vehicle feature set; the first vehicle feature set is a vehicle feature set whose similarity with the target vehicle feature set corresponding to the target vehicle is greater than or equal to a second preset similarity threshold.

5. A motion trajectory determination device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 3 is implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Citation Information

Patent Citations

  • Pedestrian recognition method and device

    CN109753920A

  • Methods and systems for video surveillance

    US20180173965A1

  • Face recognition method and apparatus, and access control attendance machine

    WO2020000196A1