A personnel positioning method and system based on a target MAC address and surveillance video

Abstract

The present invention discloses a method for positioning a person based on a target MAC address and a surveillance video. The method includes: importing target MAC address information, querying and generating a target MAC trajectory, retrieving corresponding surveillance video data according to the target MAC trajectory, generating multiple video trajectories according to the surveillance video data, determining a target video trajectory from the multiple video trajectories through collision, and binding the surveillance video data corresponding to the target video trajectory with the target MAC address information. The present invention combines the known target MAC address information with the surveillance video image recognition technology. Through big data analysis technology, it comprehensively considers the similarity between the MAC duration trajectory and the video duration trajectory as well as the similarity between the MAC count trajectory and the video count trajectory, can quickly and effectively obtain an unknown target person, greatly facilitates the investigators to determine the effective identity information of the suspected target, and improves the effective utilization rate of the research and judgment tools.

Description

Technical Field

[0001] The present invention relates to the fields of video surveillance and intelligent security, and particularly relates to a method for positioning personnel based on a target MAC address and surveillance video. Background Art

[0002] Currently, both face recognition and target MAC acquisition technologies on the market are very mature. Relevant national departments have deployed video surveillance devices and target MAC acquisition devices in the streets and alleys of cities to safeguard urban security. However, when positioning some special personnel, especially those whose identities have not been confirmed and cannot be directly recognized through video, it is often necessary to first use manual methods to analyze the trajectory of the target MAC information, and then screen through video recordings, which consumes a huge amount of manpower and has a certain impact on the detection efficiency of cases. Summary of the Invention

[0003] Aiming at the deficiencies in the prior art, the present invention provides a method for positioning personnel based on a target MAC address and surveillance video, including the following:

[0004] Import the target MAC address information, query and generate a target MAC trajectory, where the target MAC trajectory includes a target MAC duration trajectory and a target MAC frequency trajectory;

[0005] Retrieve the corresponding surveillance video data according to the target MAC trajectory, and generate multiple video trajectories according to the surveillance video data, where the video trajectory includes a video duration trajectory and a video frequency trajectory;

[0006] Associate the target MAC trajectory and the video trajectory through collision, and bind the surveillance video data and the target MAC address information.

[0007] Preferably, the query and generation of the target MAC trajectory includes:

[0008] Query the target MAC address information in multiple specified areas;

[0009] Obtain the number of occurrences of the target MAC address information in each specified area, as well as the start time and end time of each occurrence;

[0010] Generate a target MAC trajectory according to the number of occurrences of the target MAC address information in each specified area, as well as the start time and end time of each occurrence.

[0011] Preferably, the generation of the target MAC trajectory according to the number of occurrences of the target MAC address information in each specified area, as well as the start time and end time of each occurrence includes:

[0012] Obtain the number of occurrences of the target MAC address information in each specified area, and mark the number of occurrences as R i , i = 1, 2,... n, i represents the area, and obtain the target MAC count trajectory sequence {R1, R2, R3,..., R n};

[0013] Obtain the start time and end time of each occurrence of the target MAC address information in each specified area, calculate the duration and interval duration of each occurrence, and record the ratio of the average value of all duration times and the average value of all interval times in each area as S i = g i / h i , and obtain the target MAC duration trajectory {S1, S2, S3,..., S n}.

[0014] Preferably, retrieve the corresponding monitoring video data according to the target MAC trajectory, and generate multiple video trajectory sequences based on the monitoring video data, including:

[0015] Retrieve the monitoring video data of multiple specified areas corresponding to the target MAC trajectory;

[0016] Identify the faces appearing in the monitoring video data through image recognition technology, and obtain the number of occurrences of each face and the start time and end time of each occurrence;

[0017] Generate the video trajectory of each face according to the number of occurrences of each face and the start time and end time of each occurrence.

[0018] Preferably, generating the video trajectory of each face according to the number of occurrences of each face in each specified area and the time of each occurrence includes:

[0019] Obtain the number of occurrences of any face in each specified area, and mark the number of occurrences as F i , and obtain the video count sequence {F1, F2, F3,..., F n}, where the number of occurrences includes the determined number of occurrences and the suspected number of occurrences;

[0020] Obtain the start time and end time of each occurrence of this face in each specified area, calculate the duration and interval duration of each occurrence, and record the ratio of the average value of all duration times and the average value of all interval times in each area as Q i = G i / H i , and obtain the video duration trajectory {Q1, Q2, Q3,..., Q n};

[0021] Repeat the above steps until the video trajectory of each face is obtained.

[0022] Preferably, the F i Specifically:

[0023] F i = P ij + P ij-im / 2

[0024] Wherein, P ij represents the determined appearance times of this face in the i-th area, and the P ij-im represents the suspected appearance times of this face in an area.

[0025] Preferably, obtaining the start time and end time of each appearance of this face in each specified area, and calculating the duration and interval duration of each appearance includes:

[0026] If this appearance is a suspected appearance, record the duration of this appearance as 1 / 2 times the actual duration; if this appearance is a determined appearance, record the duration of this appearance as the actual duration, wherein the actual duration is the difference between the start time and end time of this appearance;

[0027] If both this appearance and the next appearance are suspected appearances, record the interval duration between these two as 1 / 4 times the actual interval duration; if either this appearance or the next appearance is a suspected appearance, record the interval duration between these two as 1 / 2 times the actual interval duration; if both this appearance and the next appearance are determined appearances, record the interval duration between these two as the actual interval duration, wherein the actual interval duration is the difference between the start time of the next appearance and the end time of this appearance.

[0028] Preferably, determining the target video trajectory from multiple video trajectories through collision, associating the target MAC trajectory and the target video trajectory, and binding the monitoring video data corresponding to the target video trajectory and the target MAC address information includes:

[0029] Calculating the similarity between the target MAC trajectory and multiple video trajectories;

[0030] Taking the video trajectory with the highest similarity as the target video trajectory, associating the target MAC trajectory with the target video trajectory, and binding the target MAC address information corresponding to the target MAC trajectory and the monitoring video data corresponding to the target video trajectory.

[0031] Preferably, calculating the similarity between the target MAC trajectory and multiple video trajectories includes:

[0032] Obtain a video track, and calculate the first similarity between the target MAC count track and the video count track in this video track, and the second similarity between the target MAC duration track and the video duration track in this video track respectively;

[0033] Calculate the similarity between the target MAC track and this video track according to the first similarity and the second similarity;

[0034] Repeat the above steps until the similarities between the target MAC track and all video tracks are obtained.

[0035] Preferably, calculating the similarity between the target MAC track and this video track according to the first similarity and the second similarity includes:

[0036] X = λX1+(1 - λ)X2

[0037] Wherein, X is the similarity, X1 is the first similarity, X2 is the second similarity, and λ is a constant.

[0038] The beneficial effects of the present invention are reflected in that: the present invention combines the known target MAC information with the monitoring video image recognition technology, and through the big data analysis technology, comprehensively considers the similarity between the MAC duration track and the video duration track and the similarity between the MAC count track and the video count track, can quickly and effectively obtain unknown target persons, greatly facilitates the investigators to determine the effective identity information of the suspected targets, and improves the effective utilization rate of the research and judgment tools. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts do not necessarily draw according to the actual scale.

[0040] Figure 1 It is a flowchart of a method for locating a person based on a target MAC address and a monitoring video;

[0041] Figure 2 It is a schematic structural diagram of a system for locating a person based on a target MAC address and a monitoring video. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] The following will describe in detail the embodiments of the technical solutions of the present invention with reference to the drawings. The following embodiments are only used to illustrate the technical solutions of the present invention more clearly, so they are only examples and cannot be used to limit the protection scope of the present invention.

[0043] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application shall have the ordinary meanings understood by those skilled in the art to which this invention pertains.

[0044] As Figure 1 shown, an embodiment of the present invention provides a method for locating a person based on a target MAC address and a surveillance video, including the following steps:

[0045] Step 1: Import the target MAC address information, query and generate a target MAC trajectory, where the target MAC trajectory includes a target MAC duration trajectory and a target MAC frequency trajectory;

[0046] In the embodiment of the present invention, the querying and generating the target MAC trajectory includes: querying the target MAC address information in multiple specified areas; obtaining the number of occurrences of the target MAC address information in each specified area and the start time and end time of each occurrence; generating a target MAC trajectory according to the number of occurrences of the target MAC address information in each specified area and the start time and end time of each occurrence.

[0047] In the embodiment of the present invention, the generating a target MAC trajectory according to the number of occurrences of the target MAC address information in each specified area and the start time and end time of each occurrence includes:

[0048] Mark the number of occurrences as R i , i = 1, 2,... n, i represents the area, to obtain a target MAC frequency trajectory sequence {R1, R2, R3,..., R n};

[0049] Calculate the duration and interval duration of each occurrence according to the start time and end time of each occurrence of the target MAC address information in each specified area, and record the ratio of the average value of all duration times and the average value of all interval times in each area as S i = g i / h i , to obtain a target MAC duration trajectory {S1, S2, S3,..., S n}.

[0050] Specifically, if the specified area is ABCD, which are represented by 1234 respectively; in area A, the number of times the target MAC address information appears is 3. The start time of the first appearance is 3:14, and the end time is 3:19; the start time of the second appearance is 3:40, and the end time is 3:46; the start time of the third appearance is 5:11, and the end time is 5:21; in area B, the number of appearances is 2. The start time of the first appearance is 4:12, and the end time is 4:19; the start time of the second appearance is 5:27, and the end time is 5:36; in area C, the number of appearances is 1. The start time of this appearance is 6:01, and the end time is 6:09; in area D, the number of appearances is 1. The start time of this appearance is 6:30, and the end time is 6:36;

[0051] Then, the target MAC count trajectory is {3, 2, 4, 1}, and the target MAC duration trajectory is {7 / 23, 8 / 68, 8, 6}.

[0052] Step 2: Retrieve the corresponding monitored video data according to the target MAC trajectory, and generate multiple video trajectories according to the monitored video data, where the video trajectories include video duration trajectories and video count trajectories;

[0053] In the embodiment of the present invention, retrieving the corresponding monitored video data according to the target MAC trajectory and generating multiple video trajectory sequences according to the monitored video data includes: retrieving the monitored video data of multiple specified areas according to the target MAC trajectory; identifying the faces appearing in the monitored video data through image recognition technology, and obtaining the number of appearances of each face and the start time and end time of each appearance; generating the video trajectory of each face according to the number of appearances of each face and the start time and end time of each appearance.

[0054] Specifically, identifying the faces appearing in the monitored video data through image recognition technology, specifically, face comparison technology can be used for duplicate checking and comparison technology. If the similarity exceeds the first threshold, it is judged as the same face; if the similarity is lower than the first threshold and higher than the second threshold, it is judged as a suspected face; if the similarity is lower than the second threshold, it is judged as a different face. In other embodiments of the invention, existing image recognition models can be used for recognition, which is not limited in this application.

[0055] Of course, in other embodiments, using image recognition technology to identify the target in the monitored data includes, but is not limited to, faces, such as vehicles.

[0056] In the embodiment of the present invention, generating the video trajectory of each face according to the number of appearances of each face in each specified area and the time of each appearance includes:

[0057] Obtain the number of occurrences of any face in each specified area, and mark the number of occurrences as F i , to obtain a video number sequence {F1, F2, F3,..., F n}, where the number of occurrences includes a determined number of occurrences and a suspected number of occurrences;

[0058] Obtain the start time and end time of each occurrence of this face in each specified area, calculate the duration and interval duration of each occurrence, calculate the average value of all duration and all interval durations in each area, and record the ratio of the average value of all duration and the average value of all interval durations in each area as Q i = G i / H i , to obtain a video duration trajectory {Q1, Q2, Q3,..., Q n};

[0059] Repeat the above steps until the video trajectory of each face is obtained.

[0060] The method for obtaining the video trajectory adopted in the embodiments of the present invention and the method for obtaining the MAC trajectory are based on the same inventive concept. Therefore, for more specific examples of obtaining the video trajectory, no further description will be given here.

[0061] It should be noted that since the MAC address information is determined data, and the faces in the surveillance video include determined faces and uncertain faces, therefore, in the embodiments of the present invention, the following methods also need to be adopted to process the number of occurrences of faces and the start time and end time of each occurrence to ensure that the obtained data is more accurate.

[0062] Specifically, the F i Specifically:

[0063] F i = P ij + P ij-im / 2

[0064] Wherein, P ij represents the determined number of occurrences of this face in the i-th area, and the P ij-im represents the suspected number of occurrences of this face in an area.

[0065] Specifically, obtaining the start time and end time of each occurrence of this face in each specified area and calculating the duration and interval duration of each occurrence includes:

[0066] If this occurrence is a suspected occurrence, record the duration of this occurrence as 1 / 2 times the actual duration; if this occurrence is a determined occurrence, record the duration of this occurrence as the actual duration, where the actual duration is the difference between the start time and end time of this occurrence;

[0067] If both the current occurrence and the next occurrence are suspected occurrences, record the time interval between these two occurrences as 1 / 4 times the actual time interval; if either the current occurrence or the next occurrence is a suspected occurrence, record the time interval between these two occurrences as 1 / 2 times the actual time interval; if both the current occurrence and the next occurrence are definite occurrences, record the time interval between these two occurrences as the actual time interval, where the actual time interval is the difference between the start time of the next occurrence and the end time of the current occurrence.

[0068] In a specific embodiment, further, the duration and time interval of each occurrence can be processed. Since data with too short a duration has little significance, it is necessary to delete data with a duration less than the preset duration threshold, and then calculate the average value of the remaining durations; since data with too long an interval has little significance, it is necessary to delete data with an interval exceeding the preset interval threshold, and then calculate the average value of the remaining intervals.

[0069] Step 3: Determine the target video track from multiple video tracks through collision, associate the target MAC track with the target video track, and bind the monitoring video data corresponding to the target video track and the target MAC address information.

[0070] Step 3 specifically includes: calculating the similarity between the target MAC track and multiple video tracks; taking the video track with the highest similarity as the target video track, associating the target MAC track with the target video track, and binding the target MAC address information corresponding to the target MAC track and the monitoring video data corresponding to the target video track.

[0071] In the embodiment of the present invention, calculating the similarity between the target MAC track and multiple video tracks includes:

[0072] Obtain a video track, and calculate the first similarity between the target MAC count track and the video count track in this video track and the second similarity between the target MAC duration track and the video duration track in this video track respectively;

[0073] Calculate the similarity between the target MAC track and this video track according to the first similarity and the second similarity;

[0074] Repeat the above steps until the similarities between the target MAC track and all video tracks are obtained.

[0075] It should be noted that both the first similarity and the second similarity are calculated as cosine similarities, and the calculation formula is:

[0076]

[0077] where xi Represents an element in the MAC trajectory, y i Represents an element in the video trajectory.

[0078] In a specific embodiment, the target MAC trajectory and the video trajectory can be normalized and then the first similarity and the second similarity are calculated respectively to reduce the amount of calculation.

[0079] In an embodiment of the present invention, calculating the similarity between the target MAC trajectory and the video trajectory according to the first similarity and the second similarity includes:

[0080] X = λX1+(1 - λ)X2

[0081] Wherein, X is the similarity, X1 is the first similarity, X2 is the second similarity, and λ is a constant, which is calculated according to experience.

[0082] Through the above method, all MAC address information can be matched with its corresponding monitored video data.

[0083] In summary, the embodiment of the present invention provides a method for locating a person based on a target MAC address and a monitored video, which combines the known target MAC information with the monitored video image recognition technology. Through big data analysis technology, it comprehensively considers the similarity between the MAC duration trajectory and the video duration trajectory and the similarity between the MAC count trajectory and the video count trajectory, and can quickly and effectively obtain unknown target persons, greatly facilitating the investigators to determine the effective identity information of the suspected target and improving the effective utilization rate of the research and judgment tools.

[0084] Such as Figure 2 As shown, based on the above inventive concept, the embodiment of the present invention further provides a system for locating a person based on a target MAC address and a monitored video. The system includes a research and judgment platform, a plurality of probes installed in designated areas for collecting MAC address information, and cameras for collecting monitored video data. Among them, the research and judgment platform is used to execute the method described in the above embodiment.

[0085] First, input the determined target MAC address information into the platform. The system generates the movement trajectory of the target MAC on the map according to the collected MAC information. The user selects the movement trajectory within a certain time period or a certain area and automatically performs data fusion and collision with the image information collected by the video monitoring. The system judges according to the principle that the image information collected by all cameras within this movement trajectory highly conforms to the target characteristics, and outputs the person image information or the vehicle image information driven by the corresponding target MAC address, etc.

[0086] For example, it is known that as Figure 2The MAC address information of the little black man and the equipment on the little black car shown are MAC1 and MAC2 respectively. When the MAC2 address information and the video image information of the little black man appear in both Area A and Area D, we can determine that the little black man is the person corresponding to the MAC2 address, and the identity of the person can be confirmed through the video image information. Similarly, the little black cars in Area A and Area C can also determine their video image information in this way, so as to obtain identities such as vehicle models, colors, and license plate numbers, and then confirm the vehicle owner information and vehicle trajectory. When there are more output results, the comparison results can be further narrowed by expanding the research and judgment area and time, and finally a one-to-one binding result can be achieved, so as to achieve the effect of personnel identity positioning.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered by the scope of the claims and the specification of the present invention.

Claims

1. A method for personnel positioning based on the target MAC address and surveillance video, characterized in that It includes the following contents: Import the target MAC address information, query and generate the target MAC trajectory, where the target MAC trajectory includes the target MAC duration trajectory and the target MAC frequency trajectory; Retrieve the corresponding monitored video data according to the target MAC trajectory, and generate multiple video trajectories according to the monitored video data, where the video trajectory includes the video duration trajectory and the video frequency trajectory; Generate the video trajectory of each face according to the appearance frequency of each face and the start time and end time of each appearance; The appearance frequency includes the determined appearance frequency and the suspected appearance frequency; Calculate the duration and interval duration of each appearance according to the start time and end time of each appearance of the face in each specified area; Determine the target video trajectory from multiple video trajectories through collision, associate the target MAC trajectory and the target video trajectory, and bind the monitored video data corresponding to the target video trajectory and the target MAC address information; The calculating the duration and interval duration of each appearance according to the start time and end time of each appearance of the face in each specified area includes: If this appearance is a suspected appearance, record the duration of this appearance as 1 / 2 times of the actual duration; if this appearance is a determined appearance, record the duration of this appearance as the actual duration, where the actual duration is the difference between the start time and end time of this appearance; If both this appearance and the next appearance are suspected appearances, record the interval duration between the two as 1 / 4 times of the actual interval duration; if this appearance or the next appearance is a suspected appearance, record the interval duration between the two as 1 / 2 times of the actual interval duration; if both this appearance and the next appearance are determined appearances, record the interval duration between the two as the actual interval duration, where the actual interval duration is the difference between the start time of the next appearance and the end time of this appearance; process the duration and interval duration of each appearance, delete the data with a duration less than the preset duration threshold, and then calculate the average value of the remaining durations.

2. The method for positioning a person based on a target MAC address and a surveillance video according to claim 1, wherein, The querying and generating the target MAC trajectory includes: Query the target MAC address information in multiple specified areas; Obtain the appearance frequency of the target MAC address information in each specified area and the start time and end time of each appearance; Generate the target MAC trajectory according to the appearance frequency of the target MAC address information in each specified area and the start time and end time of each appearance.

3. The method for positioning a person based on a target MAC address and a surveillance video according to claim 2, wherein The generating the target MAC trajectory according to the appearance frequency of the target MAC address information in each specified area and the start time and end time of each appearance includes: Mark the number of occurrences as R according to the number of occurrences of the target MAC address information in each specified area i , where i = 1, 2, …, n, and i represents the area, to obtain the target MAC number of occurrences trajectory sequence {R1, R2, R3, …, R n}; Calculate the duration and interval duration of each occurrence based on the start time and end time of each occurrence of the target MAC address information in each specified area, and record the ratio of the average value of all duration times and the average value of all interval times in each area as S i =g i / h i to obtain the target MAC duration trajectory {S1, S2, S3, …, S n}.

4. A method for positioning a person based on a target MAC address and a surveillance video according to claim 1, characterized in that, Retrieving the corresponding monitored video data according to the target MAC trajectory, and generating multiple video trajectories according to the monitored video data includes: Retrieve the monitored video data of multiple specified areas corresponding to the target MAC trajectory; Identify the faces appearing in the monitored video data through image recognition technology, and obtain the appearance frequency of each face and the start time and end time of each appearance; Generate the video track of each face according to the appearance times of each face and the start time and end time of each appearance.

5. The personnel positioning method based on the target MAC address and the monitoring video according to claim 4, wherein, The generating the video track of each face according to the appearance times of each face in each specified area and the time of each appearance includes: Mark the number of occurrences as F according to the number of occurrences of any face in each specified area i , to obtain a video number sequence {F1, F2, F3,..., F n}, where the number of occurrences includes the determined number of occurrences and the suspected number of occurrences; According to the start time and end time of each appearance of the face in each specified area, calculate the duration and interval duration of each appearance, and record the ratio of the average value of all the duration times and the average value of all the interval times in each area as Q i =G i / H i , to obtain the video duration trajectory {Q1, Q2, Q3,..., Q n}; Repeat the above steps until the video track of each face is obtained.

6. The method for positioning a person based on a target MAC address and a surveillance video according to claim 5, wherein The said F i Specifically: F i =P ij +P ij-im / 2 Among them, P ij represents the determined occurrence times of the face in the i-th area, and the P ij-im represents the suspected occurrence times of the face in an area.

7. A method for positioning a person based on a target MAC address and a surveillance video according to claim 1, characterized in that The determining the target video track from multiple video tracks through collision, associating the target MAC track with the target video track, and binding the monitoring video data corresponding to the target video track and the target MAC address information includes: Calculate the similarity between the target MAC track and multiple video tracks; Take the video track with the highest similarity as the target video track, associate the target MAC track with the target video track, and bind the target MAC address information corresponding to the target MAC track and the monitoring video data corresponding to the target video track.

8. A method for positioning a person based on a target MAC address and a surveillance video according to claim 7, characterized in that, The calculating the similarity between the target MAC track and multiple video tracks includes: Obtain a video track, and calculate the first similarity between the target MAC count track and the video count track in this video track and the second similarity between the target MAC duration track and the video duration track in this video track respectively; Calculate the similarity between the target MAC track and this video track according to the first similarity and the second similarity; Repeat the above steps until the similarities between the target MAC track and all video tracks are obtained.

9. The method for positioning a person based on a target MAC address and a surveillance video according to claim 8, wherein, The calculating the similarity between the target MAC track and this video track according to the first similarity and the second similarity includes: X = λX1+(1 - λ)X2 where X is the similarity, X1 is the first similarity, X2 is the second similarity, and λ is a constant.

Images