A trajectory classification method, an intelligent terminal and a storage medium

Abstract

This application discloses a trajectory classification method, a smart terminal, and a computer-readable storage medium. The trajectory classification method includes: acquiring a first trajectory set; wherein each trajectory in the first trajectory set is determined by image information corresponding to monitoring images from multiple monitoring points; determining a target trajectory in the first trajectory set, and determining a set of associated trajectories related to the multiple monitoring points corresponding to the target trajectory; and classifying the multiple trajectories in the set of associated trajectories. By utilizing multiple monitoring points associated with the target trajectory to determine the set of associated trajectories and classifying the multiple trajectories in the set of associated trajectories, the efficiency of associated trajectory selection and the accuracy of associated trajectory classification can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of trajectory classification, and in particular to a trajectory classification method, an intelligent terminal, and a computer readable storage medium. BACKGROUND

[0002] Video monitoring devices in cities collect massive portrait picture video data every day. Organizing and utilizing the information in the data in the form of aggregation can improve the stability and prosperity of society. SUMMARY

[0003] To solve the above problems, the present application provides a trajectory classification method, an intelligent terminal, and a computer readable storage medium, which can improve the efficiency of selecting associated trajectories and the accuracy of classifying associated trajectories.

[0004] One technical solution adopted by the present application is: a trajectory classification method, the method comprising: obtaining a first trajectory set; wherein each trajectory in the first trajectory set is determined by image information corresponding to monitoring images of multiple monitoring points; in the first trajectory set, determining a target trajectory, and determining an associated trajectory set associated with the multiple monitoring points corresponding to the target trajectory; and classifying the multiple trajectories in the associated trajectory set.

[0005] Optionally, the image information comprises face information, monitoring point information, and time information; obtaining the first trajectory set comprises: obtaining multiple monitoring image sets; wherein each monitoring image set comprises at least two monitoring images, the at least two monitoring images are collected by multiple monitoring points, and each two monitoring images correspond to the same face information; and drawing a trajectory according to the image information of the monitoring images in each monitoring image set; and collecting the trajectories of the multiple monitoring image sets to obtain the first trajectory set.

[0006] Optionally, drawing a trajectory according to the image information of the monitoring images in each monitoring image set comprises: determining a first monitoring point set corresponding to the monitoring images in each monitoring image set according to the monitoring point information; and drawing a connection in time sequence according to the time information corresponding to each monitoring point in the first monitoring point set to obtain the trajectory.

[0007] Optionally, the image information further comprises label information; and drawing a trajectory according to the image information of the monitoring images in each monitoring image set further comprises: marking a label at a position corresponding to the trajectory according to the label information and the monitoring point information.

[0008] Optionally, in the first trajectory set, the target trajectory is determined, and the associated trajectory associated with the plurality of monitoring points corresponding to the target trajectory is determined, including: determining the target trajectory according to the label on the trajectory; determining the second type of monitoring point set according to the monitoring point information corresponding to the first trajectory set, and determining the third type of monitoring point set according to the monitoring point information corresponding to the target trajectory; determining the associated monitoring point associated with the third type of monitoring point set in the second type of monitoring point set; and determining the associated trajectory in the first trajectory set according to the associated monitoring point.

[0009] Optionally, the associated monitoring point associated with the third type of monitoring point set in the second type of monitoring point set is determined, including: determining the first preset number of monitoring points closest to each monitoring point in the third type of monitoring point set in the second type of monitoring point set; and collecting the first preset number of monitoring points and the monitoring points in the third type of monitoring point set to obtain the associated monitoring point; and determining the associated trajectory in the first trajectory set according to the associated monitoring point, including: determining the trajectory including at least one associated monitoring point in the first trajectory set as the associated trajectory.

[0010] Optionally, the plurality of trajectories in the associated trajectory set is classified, including: determining the trajectory similarity of the plurality of trajectories in the associated trajectory set and the target trajectory; and classifying the plurality of trajectories in the associated trajectory set and the target trajectory according to the trajectory similarity.

[0011] Optionally, the trajectory similarity of the plurality of trajectories in the associated trajectory set and the target trajectory is determined, including: determining the first monitoring point farthest from each monitoring point in the third type of monitoring point set in the first preset number of monitoring points; determining the first distance between each monitoring point in the third type of monitoring point set and the corresponding first monitoring point; determining the upper limit of the trajectory similarity according to the sum of the similarity contributions of each monitoring point in the third type of monitoring point set and the corresponding first monitoring point based on the first distance; respectively determining the second monitoring point closest to each monitoring point in the third type of monitoring point set in the plurality of trajectories within the first distance; determining the second distance between each monitoring point in the third type of monitoring point set and the corresponding second monitoring point; and determining the lower limit of the trajectory similarity according to the sum of the similarity contributions of each monitoring point in the third type of monitoring point set and the corresponding second monitoring point based on the second distance.

[0012] Optionally, the plurality of trajectories in the associated trajectory set and the target trajectory are classified according to the trajectory similarity, including: arranging the plurality of trajectories in the associated trajectory set in a first similarity column sequence from large to small according to the lower limit of the trajectory similarity; determining a second similarity column sequence in which the lower limit of the trajectory similarity is greater than the upper limit of the trajectory similarity in the first similarity column sequence; and classifying the second preset number of trajectories and the target trajectory into a category according to the lower limit of the trajectory similarity from large to small in the second similarity column sequence.

[0013] Optionally, the method further comprises: training a trajectory classification model by taking the second preset number of trajectories as training samples; obtaining a second trajectory set; and inputting the trajectories in the second trajectory set into the trajectory recognition model to classify the trajectories in the second trajectory set.

[0014] Another technical solution adopted by the present application is to provide an intelligent terminal, which comprises a processor and a memory connected to the processor; wherein the memory stores program data, and the processor calls the program data stored in the memory to execute the trajectory classification method as described above.

[0015] Another technical solution adopted by the present application is to provide a computer readable storage medium, which stores program data, and the program data, when executed by a processor, is used to implement the trajectory classification method as described above.

[0016] The trajectory classification method provided by the present application comprises: obtaining a first trajectory set; wherein each trajectory in the first trajectory set is determined by image information corresponding to a plurality of monitoring images of monitoring points; in the first trajectory set, a target trajectory is determined, and an associated trajectory set associated with a plurality of monitoring points corresponding to the target trajectory is determined; and a plurality of trajectories in the associated trajectory set are classified. Through the above-mentioned manner, on the one hand, a plurality of behavior trajectory images can be drawn through a plurality of monitoring points and their image information corresponding to a plurality of monitoring images. On the other hand, the associated trajectory set of the target trajectory is determined through a plurality of monitoring points corresponding to the target trajectory, and a plurality of trajectories in the associated trajectory set are classified, thereby improving the efficiency of selecting the associated trajectory and the accuracy of classifying the associated trajectory. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Among them:

[0018] Figure 1 is a flowchart of the first embodiment of the trajectory classification method provided by the present application;

[0019] Figure 2 is a flowchart of an embodiment of step 21 of obtaining the first trajectory set of the present application;

[0020] Figure 3 is a flowchart of an embodiment of step 212 of drawing the trajectory of the present application;

[0021] Figure 4is a flowchart of an embodiment of step 22 of the present application;

[0022] Figure 5 is a flowchart of an embodiment of step 223 of the present application;

[0023] Figure 6 is a schematic diagram of an embodiment of determining the associated monitoring point associated with the third set of monitoring points;

[0024] Figure 7 is a flowchart of an embodiment of step 23 of the present application;

[0025] Figure 8 is a schematic diagram of an embodiment of the present application calculating the Euclidean distance of each monitoring point and the target trajectory;

[0026] Figure 9 is a schematic diagram of an embodiment of the present application calculating the sum of the similarity contribution of the associated monitoring point and the target trajectory;

[0027] Figure 10 is a flowchart of an embodiment of step 231 of the present application;

[0028] Figure 11 is a schematic diagram of another embodiment of the present application calculating the sum of the similarity contribution of the associated monitoring point and the target trajectory;

[0029] Figure 12 is a flowchart of an embodiment of step 232 of the present application;

[0030] Figure 13 is a flowchart of a third embodiment of the trajectory classification method provided by the present application;

[0031] Figure 14 is a structural schematic diagram of an intelligent terminal provided by the present application;

[0032] Figure 15 is a structural schematic diagram of an embodiment of a computer readable storage medium provided by the present application. DETAILED DESCRIPTION

[0033] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. It can be understood that the specific embodiments described herein are only used to explain the present application, rather than limit the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings, rather than all the structures. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0034] Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another. As used herein, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or". That is, unless specified otherwise, or clear from context, "X employs A or B" is intended to mean any of the natural inclusive permutations. That is, if X employs A; X employs B; or X employs both A and B, then "X employs A or B" is satisfied. In addition, the articles "a" and "an" as used in this application and the appended claims should generally be construed to mean "one or more" unless specified otherwise or clear from context to be directed to a singular form.

[0035] The steps in the embodiments of the present application are not necessarily processed in the order described, and the steps in the embodiments of the present application can be selectively rearranged, deleted, or added according to requirements. The step description in the embodiments of the present application is only an optional sequence combination, and does not represent all sequence combinations of the embodiments of the present application. The sequence of steps in the embodiments cannot be considered as a limitation of the present application.

[0036] The term "and / or" in the embodiments of the present application means any and all possible combinations of the associated listed items. It should also be noted that: when used in the specification, "includes / contains" specifies the existence of the stated features, integers, steps, operations, elements and / or components, but does not exclude the existence or addition of one or more other features, integers, steps, operations, elements and / or groups.

[0037] The terms "first", "second", and the like in the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units not listed, or optionally also includes other steps or units inherent to the process, method, product or device.

[0038] In addition, although the terms "first", "second", and the like are used repeatedly in the present application to describe various thresholds (or various elements or various applications or various instructions or various operations), etc., these thresholds (or elements or applications or instructions or operations) should not be limited by these terms. These terms are only used to distinguish one threshold (or element or application or instruction or operation) from another threshold (or element or application or instruction or operation). For example, a first trajectory set can be referred to as a second trajectory set, and a second trajectory set can also be referred to as a first trajectory set, only the ranges included by the two are different, without departing from the scope of the present application, and the first trajectory set and the second trajectory set are both sets of various image trajectories, only the two are not the same set of image trajectories.

[0039] The intelligent terminal (e.g., mobile terminal) of the embodiments of the present application can be implemented in various forms. Among them, the intelligent terminal can be a mobile terminal capable of storing image information and being accessed or sent image information, including devices such as image acquisition and identification devices (e.g., cameras and video recorders), mobile phones, smart phones, notebook computers, personal digital assistants (PDAs), tablet computers (PADs), and the like. The intelligent terminal can also be a fixed terminal capable of storing image information and being accessed or sent image information, such as digital broadcast transmitters, digital TVs, desktop computers, and the like. In the following, it is assumed that the terminal is a mobile terminal. However, those skilled in the art will understand that the configuration according to the embodiments of the present application can also be applied to terminals of the fixed type, except for elements specifically used for mobile purposes.

[0040] Referring to Figure 1 , Figure 1 is a flowchart of a first embodiment of a trajectory classification method provided by the present application. The method comprises:

[0041] Step 11: Obtain a first trajectory set.

[0042] Among them, each trajectory in the first trajectory set is determined by image information corresponding to the monitoring images of the multiple monitoring points.

[0043] Specifically, the first trajectory set is read from a database, and the first trajectory set includes multiple trajectories. Each trajectory is connected by the positions of the multiple monitoring points. The positions of the multiple monitoring points and the connection order of each trajectory can be determined from the image information corresponding to the monitoring images taken by the multiple monitoring points. The content of the monitoring image can be a portrait picture, which is obtained by intercepting part of the video data or a face camera.

[0044] Optionally, the position of the monitoring point and the shooting time of the monitoring image can be obtained from the image information corresponding to the monitoring image. According to the positions corresponding to each monitoring image and the shooting time of each monitoring image, each trajectory curve in the first trajectory set can be drawn. That is, each trajectory in the first trajectory set is a behavior trajectory image drawn by the monitoring point position and the shooting time corresponding to the monitoring image.

[0045] Step 12: In the first trajectory set, determine a target trajectory, and determine an associated trajectory set associated with the multiple monitoring points corresponding to the target trajectory.

[0046] Specifically, in the first trajectory set, a target trajectory is determined. The target trajectory can be any trajectory or a trajectory with a special identifier. The associated trajectory set is a set of trajectories associated with at least one monitoring point in the target trajectory in the first trajectory set. The associated trajectory can be a trajectory that overlaps with at least one monitoring point of the target trajectory or has a distance less than a preset value.

[0047] Step 13: Classify the trajectories in the associated trajectory set.

[0048] Specifically, the trajectories in the associated trajectory set are classified according to the distance between each associated trajectory in the associated trajectory set and each monitoring point of the target trajectory.

[0049] Optionally, the classification method can be to classify the trajectories in the associated trajectory set that overlap with one monitoring point of the target trajectory, or to classify the trajectories in the associated trajectory set that have a distance less than a preset value from at least one monitoring point of the target trajectory, or to classify the trajectories in the associated trajectory set that have a distance less than a preset value from all monitoring points of the target trajectory, and the like.

[0050] Unlike the prior art, the trajectory classification method provided by the embodiment includes: obtaining a first trajectory set; each trajectory in the first trajectory set is determined by image information corresponding to monitoring images of multiple monitoring points; in the first trajectory set, a target trajectory is determined, and an associated trajectory set associated with the multiple monitoring points corresponding to the target trajectory is determined; and the trajectories in the associated trajectory set are classified. Through the above method, on the one hand, the monitoring images and the image information corresponding thereto captured by the multiple monitoring points can obtain the behavior trajectory image corresponding to the transaction in the multiple monitoring images. On the other hand, the associated trajectory set is determined using the multiple monitoring points associated with the target trajectory, and the trajectories in the associated trajectory set are classified, which can improve the efficiency of selecting the associated trajectory and the accuracy of classifying the associated trajectory.

[0051] The above optional embodiments are combined and further optimized and expanded based on the above technical solutions to obtain a second embodiment of the trajectory classification method provided by the present application, which includes:

[0052] Step 21: Obtain a first trajectory set; each trajectory in the first trajectory set is determined by image information corresponding to monitoring images of multiple monitoring points.

[0053] Specifically, the image information includes face information, monitoring point information and time information. The face information represents face features corresponding to the monitoring images captured by each monitoring point, and whether the faces in the multiple monitoring images are faces of the same person can be determined according to the face features of the monitoring images. The monitoring point information represents positions of the monitoring points corresponding to the monitoring images, and the shooting distance between the multiple monitoring images can be determined according to the positions of the monitoring images. The time information represents times corresponding to the monitoring images captured by each monitoring point, and the time sequence of the faces appearing in the multiple monitoring images corresponding to the monitoring points can be determined according to the shooting times of the monitoring images.

[0054] Referring to Figure 2 , Figure 2 is a flowchart of an embodiment of step 21 of acquiring the first trajectory set in the present application, and step 21 specifically includes the following steps.

[0055] Step 211: Acquire multiple monitoring image sets.

[0056] Each monitoring image set includes at least two monitoring images, and each two monitoring images in the monitoring image set correspond to the same face information. That is, any one of the multiple monitoring image sets represents a set of at least two monitoring images of a person.

[0057] Step 212: Draw a trajectory according to the image information of the monitoring images in each monitoring image set.

[0058] Referring to Figure 3 , Figure 3 is a flowchart of an embodiment of step 212 of drawing a trajectory in the present application, and step 212 specifically includes the following steps.

[0059] Step 2121: Determine a first monitoring point set corresponding to the monitoring images in each monitoring image set according to the monitoring point information.

[0060] Specifically, according to the monitoring point information of each monitoring image set in the multiple monitoring image sets, all monitoring points corresponding to all monitoring images in each monitoring image set are determined, and the first monitoring point set is formed by collecting all the monitoring points.

[0061] Step 2122: Draw a trajectory by connecting the monitoring points in the first monitoring point set according to the time information of the monitoring points in the first monitoring point set in time sequence.

[0062] Specifically, in each monitoring image set, the monitoring points are connected in time sequence according to the time information of the monitoring points in the first monitoring point set of the monitoring image set, so as to draw a trajectory of the monitoring image set.

[0063] Step 213: collecting the trajectories of the plurality of monitoring image sets to obtain a first trajectory set.

[0064] In another embodiment, the image information of the monitoring image further comprises label information, wherein the label information can be represented as a special mark of the monitoring image.

[0065] Therefore, according to the image information of the monitoring image in each monitoring image set, the trajectory obtained by drawing can further comprise: according to the label information and the monitoring point information, marking a label on the position corresponding to the trajectory.

[0066] For example, the monitoring image in a monitoring image set is the image of a key monitoring person. After obtaining the face information of the monitoring image, a label of “key monitoring” can be added to each monitoring image in the monitoring image set as the label information. When the trajectory of the monitoring image set is drawn, according to the label information and the monitoring point information of each monitoring image, a corresponding label mark is marked on the position of the corresponding monitoring point of the trajectory.

[0067] Step 22: determining a target trajectory in the first trajectory set, and determining an associated trajectory set associated with the plurality of monitoring points corresponding to the target trajectory.

[0068] Referring to Figure 4 , Figure 4 is a flowchart of an embodiment of step 22 of the present application, and step 22 specifically comprises the following steps:

[0069] Step 221: determining the target trajectory according to the label on the trajectory.

[0070] Step 222: determining a second type of monitoring point set according to the monitoring point information corresponding to the first trajectory set, and determining a third type of monitoring point set according to the monitoring point information corresponding to the target trajectory.

[0071] Specifically, all monitoring point information corresponding to the monitoring image in the first trajectory set is obtained, and each monitoring point corresponding to the monitoring point information is collected into the second type of monitoring point set, and all monitoring point information corresponding to the monitoring image in the target trajectory is obtained, and each monitoring point corresponding to the monitoring point information is collected into the third type of monitoring point set.

[0072] Step 223: determining an associated monitoring point in the second type of monitoring point set associated with the third type of monitoring point set.

[0073] Referring to Figure 5 , Figure 5 is a flowchart of an embodiment of step 223 of the present application, and step 223 specifically comprises the following steps:

[0074] Step 2231: determining the first preset number of monitoring points closest to each monitoring point in the third set of monitoring points in the second set of monitoring points.

[0075] Optionally, the first preset number of monitoring points can be denoted by λ, which can be 3, 4, 5, etc., which is not specifically limited here.

[0076] Step 2232: collecting the first preset number of monitoring points and the monitoring points in the third set of monitoring points to obtain associated monitoring points.

[0077] Referring to Figure 6 , Figure 6 is a schematic diagram of an embodiment of determining the associated monitoring points associated with the third set of monitoring points. Wherein, the range A is the second set of monitoring points, and the range A is the positions of all monitoring points corresponding to the monitoring images in the first set of trajectories. Points a, b, c are all monitoring points corresponding to the monitoring images in the target trajectory R, i.e. (a, b, c) is the third set of monitoring points. The first preset number of monitoring points is set to λ=3. Therefore, it is necessary to determine the three monitoring points closest to the points a, b, c in the range A, i.e. the points 1, 2, 3 closest to the point a in the range B, the points 4, 5, 6 closest to the point b in the range C, and the points 7, 8, 9 closest to the point c in the range D. Collecting points 1, 2, 3, 4, 5, 6, 7, 8, 9 and points a, b, c to obtain associated monitoring points.

[0078] Step 224: determining the associated trajectories in the first set of trajectories according to the associated monitoring points.

[0079] Specifically, in the first set of trajectories, the trajectories (except the target trajectory) including at least one associated monitoring point are determined as the associated trajectories.

[0080] Step 23: classifying the multiple trajectories in the set of associated trajectories.

[0081] Referring to Figure 7 , Figure 7 is a flowchart of an embodiment of step 23 of the present application, and step 23 specifically includes the following:

[0082] Step 231: determining the trajectory similarity of the multiple trajectories in the set of associated trajectories and the target trajectory.

[0083] In an embodiment, the trajectory similarity of the multiple trajectories and the target trajectory is calculated by using the Euclidean distance to calculate the sum of the similarity contribution values of the associated monitoring points of the multiple trajectories in the set of associated trajectories and the target trajectory.

[0084] Firstly, the Euclidean distance between the associated monitoring point of the associated trajectory and the target trajectory is calculated, which is calculated by the following formula:

[0085] .

[0086] wherein q i represents the position of the associated monitoring point of an associated trajectory Q, R represents the target trajectory, p j represents the position of each monitoring point of the target trajectory, Dist e represents the absolute value of the Euclidean distance between any two monitoring points of the associated trajectory and the target trajectory. Dist q represents the minimum value among the absolute values of each Euclidean distance.

[0087] Referring to Figure 8 , Figure 8 is a schematic diagram of the embodiment of the present application for calculating the Euclidean distance between each monitoring point and the target trajectory. Wherein R represents the target trajectory, p1-p8 represents the position of each monitoring point of the target trajectory R, q1-q3 represents the three closest associated monitoring points of an associated trajectory Q to the target trajectory R, Dist e represents the absolute value of the Euclidean distance between two monitoring points. According to the calculation formula of the Euclidean distance, the absolute value of the Euclidean distance between the associated monitoring point q1 and the closest monitoring point p6 of the target trajectory R is Dist e (q1, p6) = 1.5; the absolute value of the Euclidean distance between the associated monitoring point q2 and the closest monitoring point p4 of the target trajectory R is Dist e (q2, p4) = 0.1; the absolute value of the Euclidean distance between the associated monitoring point q3 and the closest monitoring point p7 of the target trajectory R is Dist e (q3, p7) = 0.1.

[0088] Secondly, the sum of the similarity contribution of the associated monitoring point of each associated trajectory to the target trajectory is calculated, and then the trajectory similarity between the trajectory in the associated trajectory set and the target trajectory is determined. It is calculated by the following formula:

[0089] .

[0090] Referring to Figure 9 , Figure 9 is a schematic diagram of the embodiment of the present application for calculating the sum of the similarity contribution of the associated monitoring point to the target trajectory. Wherein the curve V = e -xThis represents the similarity contribution value between the associated monitoring point and the target trajectory. Adding the similarity contribution values ​​of each associated monitoring point to the target trajectory yields the sum of the similarity contributions, which is the trajectory similarity between the associated trajectory and the target trajectory. In the previous embodiment, the trajectory similarity Q between the associated trajectory Q and the target trajectory R is Sim(Q,R)=e -1.5 +e -0.1 +e -0.1 .

[0091] In another embodiment, the similarity between multiple trajectories and the target trajectory is calculated by using Euclidean distance to calculate the upper and lower similarity limits of the associated monitoring points of multiple trajectories in the associated trajectory set and the target trajectory.

[0092] See Figure 10 , Figure 10 This is a flowchart illustrating an embodiment of step 231 of this application. Step 231 specifically includes the following:

[0093] Step 2311: Among the first preset number of monitoring points, determine the first monitoring point that is farthest from each monitoring point in the third type of monitoring point set.

[0094] Step 2312: Determine the first distance between each monitoring point in the third type of monitoring point set and the corresponding first monitoring point.

[0095] For example, in Figure 6 In the described embodiment, a first preset number of monitoring points is set to λ = 3. Among points 1, 2, and 3, the monitoring point farthest from monitoring point a on the target trajectory R is point 3, i.e., point 3 is the first monitoring point; among points 4, 5, and 6, the monitoring point farthest from monitoring point b on the target trajectory R is point 6, i.e., point 6 is the first monitoring point; and among points 7, 8, and 9, the monitoring point farthest from monitoring point c on the target trajectory R is point 9, i.e., point 9 is the first monitoring point. Next, the distances between points a and 3 (radius1, i.e., the first distance, which can be understood as the radius between the center point a and the endpoint 3) are calculated, as are the distances between points b and 6 (radius2, i.e., the first distance, which can be understood as the radius between the center point b and the endpoint 6) and the distances between points c and 9 (radius3, i.e., the first distance, which can be understood as the radius between the center point c and the endpoint 9) are calculated respectively.

[0096] Step 2313: Based on the first distance, determine the sum of the similarity contributions between each monitoring point in the third type of monitoring point set and the corresponding first monitoring point, so as to obtain the upper limit of trajectory similarity.

[0097] The upper limit of the trajectory similarity between the target trajectory R and the associated trajectory Q is only related to each monitoring point of the target trajectory R and the first distance between each monitoring point of the target trajectory R and the first monitoring point corresponding to the monitoring point. That is, when each monitoring point of the target trajectory R is determined to be a distance from the nearest first preset number of monitoring points λ, the first monitoring point farthest from each monitoring point of the target trajectory R is determined, and then the first distance between each monitoring point of the target trajectory R and the corresponding first monitoring point is determined, that is, the upper limit of the trajectory similarity between the target trajectory R and the associated trajectory Q is determined. Therefore, the upper limit of the trajectory similarity between the target trajectory R and all associated trajectories Q is the same and determined.

[0098] The upper limit of the trajectory similarity is calculated by the following formula:

[0099] .

[0100] The upper limit UB (upper bound) of the similarity of the target trajectory Q (a, b, c) is defined as the sum of the similarity contributions provided by each radius (first distance) in the trajectory Q. Therefore, in the embodiment of step 2312, the sum of the similarity contributions of each monitoring point (point a, point b, point c) of the target trajectory Q and the corresponding first monitoring point (point 3, point 6, point 9) is e -radius1 +e -radius2 +e -radius3 That is, the upper limit of the trajectory similarity between the associated trajectory Q and the target trajectory R is UB n = e -radius1 +e -radius2 +e -radius3 .

[0101] Step 2314: Within the first distance, respectively determine the second monitoring point closest to each monitoring point in the third set of monitoring points for a plurality of trajectories.

[0102] Step 2315: Determine the second distance between each monitoring point in the third set of monitoring points and the corresponding second monitoring point.

[0103] Step 2316: According to the second distance, determine the sum of the similarity contributions of each monitoring point in the third set of monitoring points and the corresponding second monitoring point to obtain the lower limit of the trajectory similarity.

[0104] Specifically, in the radius (first distance) corresponding to each monitoring point of the target trajectory R, the second monitoring point closest to each monitoring point of the target trajectory R is determined in the radius corresponding to the associated trajectory Q. Wherein, if there is no monitoring point in the radius corresponding to a monitoring point in the third set of monitoring points, the corresponding monitoring point in the third set of monitoring points has no second monitoring point closest to it. After determining each second monitoring point, the second distance between each second monitoring point and the monitoring point in the third set of monitoring points corresponding to it is determined. Then the lower limit of the trajectory similarity between each associated trajectory Q and the target trajectory R is calculated by using the second distance.

[0105] Wherein, the lower limit of the trajectory similarity between each associated trajectory Q and the target trajectory R is calculated by the following formula:

[0106]

[0107] Wherein, R x represents the associated trajectory Q, q i represents each monitoring point in the third set of monitoring points, p j i represents the associated monitoring point in the associated trajectory Q corresponding to each monitoring point in the third set of monitoring points, max represents the maximum of the similarity contribution of the monitoring point on the associated trajectory Q to the monitoring point in the third set of monitoring points (i.e. the similarity contribution of the second monitoring point closest to each monitoring point in the third set of monitoring points to the associated trajectory).

[0108] Reference Figure 11 , Figure 11 is a schematic diagram of another embodiment of the application for calculating the sum of the similarity contribution of the associated monitoring point to the target trajectory. Wherein, point q1, point q2, point q3 are each monitoring point in the third set of monitoring points of the target trajectory R; range B is the center of point q1 with the first distance as the radius; range C is the center of point q2 with the first distance as the radius; range D is the center of point q3 with the first distance as the radius; the associated trajectory Q has monitoring points p1, p2, p3, p4, p5, wherein p1 is in range B, p2 and p3 are in range C and p2 is closer to the corresponding point q2. Therefore, according to the lower limit calculation formula of the trajectory similarity between the associated trajectory Q and the target trajectory R, it can be obtained that the similarity contribution value of p1 to point q1 is e -dist(p1,q1) , the similarity contribution value of p2 to point q2 is e -dist(p2,q2) Therefore, the lower limit of the trajectory similarity between the associated trajectory Q and the target trajectory R is LB = e -dist(p1,q1) + e -dist(p2,q2) . Wherein, e -dist(p3,q2)The reason for adding the trajectory similarity of the target trajectory R to the associated trajectory Q pair is that point p2 is closer to point q2 than point p3, and therefore point p2 can provide a higher similarity contribution.

[0109] Step 232: classifying the target trajectory and the plurality of trajectories in the associated trajectory set according to the trajectory similarity.

[0110] Referring to Figure 12 , Figure 12 is a flowchart of an embodiment of step 232 of the present application, and step 232 specifically includes the following steps:

[0111] Step 2321: arranging the plurality of trajectories in the associated trajectory set in a first similarity column sequence according to the trajectory similarity lower limit from large to small.

[0112] Step 2322: determining a second similarity column sequence in the first similarity column sequence, in which the trajectory similarity lower limit is greater than the trajectory similarity upper limit.

[0113] Step 2323: in the second similarity column sequence, classifying the second preset number of trajectories and the target trajectory according to the trajectory similarity lower limit from large to small.

[0114] For example, in an embodiment, the number of associated trajectories in the associated trajectory set is N, and the N associated trajectories are arranged in a first similarity column sequence according to the trajectory similarity lower limit with the target trajectory from large to small. In the first similarity column sequence, the associated trajectories with a trajectory similarity lower limit less than the trajectory similarity upper limit are removed, and the remaining associated trajectories (the number P and P is less than or equal to N) are arranged in a second similarity column sequence according to the trajectory similarity lower limit with the target trajectory from large to small. In the second similarity column sequence, the second preset number K of associated trajectories and the target trajectory are selected to form a class of trajectories, wherein the trajectories in the class of trajectories are the most similar trajectories to the target trajectory.

[0115] If the number P of the remaining associated trajectories is less than the second preset number K of associated trajectories, return to step 2231, and expand the number λ of the first preset number of monitoring points to expand the value of the number N of the associated trajectories in the associated trajectory set, so as to satisfy that the number P is greater than or equal to the number K.

[0116] Through the above embodiment, the second preset number of track sets with the highest similarity to the target track can be found. According to the Bayesian prior probability, that is, in a track set with a high track similarity, the probability that two tracks belong to the same person is greater. For example, in the same space-time domain, the probability that two captured photos are of the same person is greater than the probability that two people in different space-time domains are the same person. The Bayesian prior probability is the same space-time domain, so for picture data in the same space-time domain, the model obtained by training the clustering algorithm is different from the model without the prior probability. Therefore, in another embodiment, the model obtained by training the algorithm according to the second preset number of track sets with the highest similarity to the target track can be used to classify the tracks.

[0117] The above optional implementation is combined, and the above technical solution is further optimized and expanded to obtain the third embodiment of the track classification method provided by the present application. The method comprises:

[0118] Referring to Figure 13 , Figure 13 is a flowchart of the third embodiment of the track classification method provided by the present application. The method comprises:

[0119] Step 31: A track classification model is trained by taking the second preset number of tracks as training samples.

[0120] Specifically, the second preset number of tracks is taken as a prior probability, and corresponding clustering algorithm training is performed thereon to obtain a model suitable for a high-similarity track set of the target track.

[0121] Step 32: A second track set is obtained.

[0122] The second track set is similar to the first track set, the second track set is read from a database, and each track in the second track set is determined by image information corresponding to a plurality of monitoring images.

[0123] Step 33: The tracks in the second track set are input into the track recognition model to classify the tracks in the second track set.

[0124] Specifically, the track recognition model calculates the upper and lower similarity limits of each track in the second track set with respect to the target track, and selects a track that meets a preset similarity threshold to be classified into a class.

[0125] Distinguish from the prior art, by the above manner, on one hand, the multiple monitoring points associated with the target trajectory are utilized to determine the associated trajectory set and classify the multiple trajectories in the associated trajectory set, which can improve the efficiency of associated trajectory selection and the accuracy of associated trajectory classification. On the other hand, the trajectory recognition model trained by the high-similarity trajectory set of the target trajectory is utilized to classify the image trajectory, which can improve the efficiency and accuracy of trajectory classification.

[0126] Reference Figure 14 , Figure 14 A structure schematic diagram of an intelligent terminal is provided in the present application, the intelligent terminal 100 comprises a processor 101 and a memory 102 connected with the processor 101, wherein the memory 102 stores program data, and the processor 101 calls the program data stored in the memory 102 to execute the trajectory classification method described above.

[0127] Optionally, in an embodiment, the processor 101 is configured to execute the program data to implement the following method: obtaining a first trajectory set; wherein each trajectory in the first trajectory set is determined by image information corresponding to monitoring images of multiple monitoring points; in the first trajectory set, a target trajectory is determined, and an associated trajectory set associated with multiple monitoring points corresponding to the target trajectory is determined; and multiple trajectories in the associated trajectory set are classified.

[0128] The processor 101 can also be referred to as a CPU (Central Processing Unit). The processor 101 can be an electronic chip with signal processing capability. The processor 101 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0129] The memory 102 can be a memory stick, a TF card, etc., and can store all information in the intelligent terminal 100, including input raw data, computer programs, intermediate running results and final running results. It stores and retrieves information according to the location specified by the processor 101. With the memory 102, the intelligent terminal 100 has a memory function and can work normally. The memory 102 of the intelligent terminal 100 can be divided into main memory (memory) and auxiliary memory (external memory) according to the purpose, and there is also a classification method of external memory and internal memory. The external memory is usually a magnetic medium or an optical disc, etc., which can store information for a long time. The memory refers to the storage component on the motherboard, which is used to store the data and programs currently being executed, but only used to temporarily store programs and data, and the data will be lost when the power is off.

[0130] In several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the above-described implementation of the intelligent terminal 100 is only illustrative. For example, the determination of the associated trajectory set according to the plurality of monitoring points associated with the target trajectory and the classification of the plurality of trajectories in the associated trajectory set, the selection manner of the second type of monitoring point set and the third type of monitoring point set, and the like are only a set manner. In actual implementation, another division manner can be used, for example, the associated trajectory set and the first similarity column sequence can be combined or can be combined into another system, or some features can be ignored or not executed.

[0131] In addition, each functional unit (such as the monitoring image database and the trajectory set database) in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0132] Referring to Figure 15 , Figure 15 The structure of an embodiment of the computer-readable storage medium provided in the present application is shown in the structural schematic diagram. The computer-readable storage medium 110 stores the program instructions 111 capable of implementing all the above methods.

[0133] If the integrated unit of each functional unit in each embodiment of the present application is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in the computer-readable storage medium 110. Based on such understanding, the technical solutions of the present application or the whole or part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer-readable storage medium 110 includes a plurality of instructions in one program instruction 111 to cause a computer device (which can be a personal computer, a system server, or a network device, etc.), an electronic device (such as MP3, MP4, etc., which can also be a mobile terminal such as a mobile phone, a tablet computer, a wearable device, etc., or a desktop computer, etc.), or a processor to execute all or part of the steps of the method of each embodiment of the present application.

[0134] Optionally, in an embodiment, the program instructions 111, when executed by the processor, are used to implement the following method: obtaining a first trajectory set; wherein each trajectory in the first trajectory set is determined by image information corresponding to the monitoring images of a plurality of monitoring points; in the first trajectory set, determining a target trajectory and determining an associated trajectory set associated with the plurality of monitoring points corresponding to the target trajectory; and classifying a plurality of trajectories in the associated trajectory set.

[0135] Those skilled in the art will appreciate that embodiments of the application can be readily used as a method, a system or a computer program product. Accordingly, the application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the application can take the form of a computer program product on one or more computer readable storage media (including, but not limited to, disk memory, CD-ROMs, optical storage devices, etc.) having computer usable program code embodied therein.

[0136] The computer program instructions 111 can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks 110. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing each of the flowchart blocks or the functions noted in the blocks

[0137] These computer readable storage media 110 can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks 110. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing each of the flowchart blocks or the functions noted in the blocks

[0138] These computer readable storage media 110 can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks 110. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing each of the flowchart blocks or the functions noted in the blocks

[0139] In an embodiment, the programmable data processing apparatus comprises a processor and a memory. The processor can also be referred to as a CPU (Central Processing Unit). The processor can be an electronic chip with the processing capability of signals. The processor can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor or the like.

[0140] The memory can be a memory stick, a TF card, etc., which stores and retrieves information according to the location specified by the processor. The memory can be classified into a main memory (internal memory) and an auxiliary memory (external memory) according to the use, or classified into an external memory and an internal memory. The external memory is usually a magnetic medium or an optical disc, etc., which can store information for a long time. The internal memory refers to a storage component on the motherboard, which is used to store data and programs currently being executed, but is only used to temporarily store programs and data, and the data will be lost when the power is turned off or disconnected.

[0141] The above description is only an embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation according to the content of the specification and drawings, or direct or indirect application in other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A trajectory classification method characterized by, The method comprises: acquiring a first trajectory set; wherein each trajectory in the first trajectory set is determined by image information corresponding to monitoring images of a plurality of monitoring points; in the first trajectory set, determining a target trajectory, and determining an associated trajectory set associated with a plurality of monitoring points corresponding to the target trajectory; classifying a plurality of trajectories in the associated trajectory set; the classifying a plurality of trajectories in the associated trajectory set comprises: in a first preset number of monitoring points, determining a first monitoring point farthest from each monitoring point in a third set of monitoring points; wherein the first preset number of monitoring points is the closest monitoring point in a second set of monitoring points to each monitoring point in the third set of monitoring points; wherein the third set of monitoring points is determined by monitoring point information corresponding to the target trajectory; the second set of monitoring points is determined according to monitoring point information corresponding to the first trajectory set; determining a first distance between each monitoring point in the third set of monitoring points and the corresponding first monitoring point; determining the sum of the similarity contribution of each monitoring point in the third set of monitoring points and the corresponding first monitoring point according to the first distance, to obtain an upper limit of the trajectory similarity; determining a second monitoring point closest to each monitoring point in the third set of monitoring points in the plurality of trajectories within the first distance; determining a second distance between each monitoring point in the third set of monitoring points and the corresponding second monitoring point; determining the sum of the similarity contribution of each monitoring point in the third set of monitoring points and the corresponding second monitoring point according to the second distance, to obtain a lower limit of the trajectory similarity; classifying the plurality of trajectories in the associated trajectory set and the target trajectory according to the upper limit and the lower limit of the trajectory similarity.

2. The trajectory classification method of claim 1, wherein: the image information comprises face information, monitoring point information and time information; the acquiring a first trajectory set comprises: acquiring a plurality of monitoring image sets; wherein each monitoring image set comprises at least two monitoring images, the at least two monitoring images are collected by the plurality of monitoring points, and each of the two monitoring images corresponds to the same face information; drawing a trajectory according to the image information of the monitoring images in each monitoring image set; pooled trajectories of the plurality of monitoring image sets to obtain the first trajectory set.

3. The trajectory classification method of claim 2, wherein: the drawing a trajectory according to the image information of the monitoring images in each monitoring image set comprises: determining a first set of monitoring points corresponding to the monitoring images in each monitoring image set according to the monitoring point information; according to the time information corresponding to each monitoring point in the first set of monitoring points, sequentially drawing connections in chronological order to obtain the trajectory.

4. The trajectory classification method of claim 3, wherein: the image information further comprises label information; The drawing of the trajectory according to the image information of the monitoring images in each of the monitoring image sets further includes: According to the label information and the monitoring point information, labels are marked at positions corresponding to the trajectory. 5.The trajectory classification method of claim 4, wherein The determination of the target trajectory and the associated trajectory associated with the monitoring points corresponding to the target trajectory in the first trajectory set includes: According to the labels on the trajectory, the target trajectory is determined; According to the monitoring point information corresponding to the first trajectory set, a second type of monitoring point set is determined, and according to the monitoring point information corresponding to the target trajectory, a third type of monitoring point set is determined; In the second type of monitoring point set, the associated monitoring points associated with the third type of monitoring point set are determined; According to the associated monitoring points, the associated trajectory is determined in the first trajectory set. 6.The trajectory classification method of claim 5, wherein The determination of the associated monitoring points associated with the third type of monitoring point set in the second type of monitoring point set includes: In the second type of monitoring point set, the first preset number of monitoring points closest to each monitoring point in the third type of monitoring point set are determined; The first preset number of monitoring points and the monitoring points in the third type of monitoring point set are collected to obtain the associated monitoring points; The determination of the associated trajectory according to the associated monitoring points in the first trajectory set includes: In the first trajectory set, the trajectory including at least one of the associated monitoring points is determined as the associated trajectory. 7.The trajectory classification method of claim 1, wherein The classification of the multiple trajectories in the associated trajectory set and the target trajectory according to the upper limit and the lower limit of the trajectory similarity includes: The multiple trajectories in the associated trajectory set are arranged in a first similarity column sequence from large to small according to the lower limit of the trajectory similarity; In the first similarity column sequence, a second similarity column sequence in which the lower limit of the trajectory similarity is greater than the upper limit of the trajectory similarity is determined; In the second similarity column sequence, the second preset number of trajectories and the target trajectory are classified into one category according to the lower limit of the trajectory similarity from large to small. 8.The trajectory classification method of claim 7, wherein The method further includes: The second preset number of trajectories are used as training samples to train a trajectory classification model; A second trajectory set is obtained; The trajectories in the second trajectory set are input into the trajectory recognition model to classify the trajectories in the second trajectory set.

9. A smart terminal, characterized by The intelligent terminal includes a processor and a memory connected to the processor, wherein the memory stores program data, and the processor calls the program data stored in the memory to execute the trajectory classification method of any one of claims 1-8.

10. A computer-readable storage medium, internally storing program instructions, characterized in that, The program instructions are executed to implement the trajectory classification method of any one of claims 1-8.

Images