Camera point data processing method and device, storage medium and electronic device

Abstract

The present application discloses a method and device for processing camera point data, a storage medium and an electronic device, wherein the method comprises: parsing objects in a set of snapshot data of a set of camera points in a preset area within a preset time to obtain a set of target objects and snapshot data of each target object in the set of target objects; fitting a movement trajectory of each target object according to the snapshot data of each target object, wherein the movement trajectory is used to indicate a camera point predicted to be passed by each target object; determining the total number of snapshot data of each target object as the actual number of snapshots of a set of camera points, and determining the total number of camera points contained in the movement trajectory of each target object as the predicted number of snapshots of a set of camera points; and performing point abnormality prediction according to the actual number of snapshots and the predicted number of snapshots to predict whether there are abnormal camera points in a set of camera points.

Description

Technical Field

[0001] The present application relates to the field of image processing, and more specifically, to a method and device for processing camera point data, a storage medium, and an electronic device. Background Art

[0002] At present, there are often thousands to tens of thousands of camera points in some designated areas (for example, a city). Due to different construction periods, different hardware and software configurations, and different camera operation times, the monitoring quality of different cameras varies. The monitoring quality of some cameras is very low due to reasons such as the height of the camera being too high or too low, the installation angle making it difficult to capture the target, and the camera being blocked by grown vegetation.

[0003] In the related art, the quality assessment of camera construction points is usually done by passing n marked targets along the route where the cameras are installed. Since the targets, the routes, and the monitoring points are known, it is not difficult to assess the quality of camera construction along a route. However, this method is too costly and the assessment time is lengthy.

[0004] It can be seen from this that the data processing method of camera points in the related art has the problem of high evaluation cost due to the large number of marked targets. Summary of the invention

[0005] The embodiments of the present application provide a camera point data processing method and device, a storage medium and an electronic device to at least solve the problem of high evaluation cost caused by a large number of marked targets in the camera point data processing method in the related art.

[0006] According to one aspect of an embodiment of the present application, a method for processing camera point data is provided, comprising: parsing objects appearing in a set of snapshot data corresponding to a set of camera points to obtain a set of target objects and snapshot data corresponding to each target object in the set of target objects, wherein the set of camera points are points where cameras are installed in a preset area, the set of snapshot data includes snapshot data of a camera at each camera point in the set of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data of each target object appearing in the set of snapshot data; fitting an image based on the snapshot data corresponding to each target object. The movement trajectory of each target object, wherein the movement trajectory of each target object is used to indicate the predicted camera points passed by each target object; the total number of snapshot data contained in the snapshot data corresponding to each target object is determined as the actual snapshot number of the group of camera points, and the total number of camera points contained in the movement trajectory of each target object is determined as the predicted snapshot number of the group of camera points; based on the actual snapshot number of the group of camera points and the predicted snapshot number of the group of camera points, point anomaly prediction is performed on the group of camera points to predict whether there are abnormal camera points in the group of camera points.

[0007] According to another aspect of the embodiments of the present application, there is also provided a data processing device for camera points, including: a parsing unit, used to parse objects appearing in a group of snapshot data corresponding to a group of camera points, to obtain a group of target objects and snapshot data corresponding to each target object in the group of target objects, wherein the group of camera points are points where cameras are installed in a preset area, the group of snapshot data includes the snapshot data of the camera at each camera point in the group of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data of each target object appearing in the group of snapshot data; a fitting unit, used to fit the snapshot data corresponding to each target object according to the snapshot data. The motion trajectory of each target object is used to indicate the predicted camera points passed by each target object; a first determining unit is used to determine the total number of captured data contained in the captured data corresponding to each target object as the actual number of captured shots of the group of camera points, and to determine the total number of camera points contained in the motion trajectory of each target object as the predicted number of captured shots of the group of camera points; a first prediction unit is used to perform point anomaly prediction on the group of camera points according to the actual number of captured shots of the group of camera points and the predicted number of captured shots of the group of camera points, so as to predict whether there are abnormal camera points in the group of camera points.

[0008] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the above-mentioned camera point data processing method when running.

[0009] According to another aspect of the embodiments of the present application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the camera point data processing method through the computer program.

[0010] In an embodiment of the present application, a method is adopted to determine the overall construction quality of camera points in a preset area according to the number of snapshot data of camera points in a preset area combined with the number of camera points involved in the actual movement trajectory of the target object corresponding to the snapshot data, and the objects appearing in a group of snapshot data corresponding to a group of camera points are parsed to obtain a group of target objects and snapshot data corresponding to each target object in the group of target objects, wherein a group of camera points are points where cameras are installed in a preset area, a group of snapshot data includes snapshot data of the camera at each camera point in a group of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data of each target object appearing in the group of snapshot data; based on the snapshot data corresponding to each target object, the movement trajectory of each target object is fitted, wherein the movement trajectory of each target object is used to indicate the predicted camera point passed by each target object; the total number of snapshot data contained in the snapshot data corresponding to each target object is determined as the actual number of snapshots of a group of camera points, and the total number of camera points contained in the movement trajectory of each target object is determined as A predicted number of snapshots of a group of camera points; based on the actual number of snapshots of a group of camera points and the predicted number of snapshots of a group of camera points, point anomaly prediction is performed on a group of camera points to predict whether there are abnormal camera points in a group of camera points. Since the correspondence between the cameras in a preset area and the corresponding camera points can be pre-configured (for example, the camera at each camera point is connected to the corresponding back-end platform system during installation and reports the corresponding location information), the target object in the preset area can be determined according to the acquired snapshot data corresponding to each camera point in the preset area, and then the movement trajectory of the target object can be determined. By comparing the number of camera points included in each movement trajectory with the number of camera points actually included in the corresponding snapshot data, the construction quality of the camera points in a preset area can be evaluated (for example, predicting whether there are abnormal camera points in a group of camera points). This can achieve the purpose of completing the evaluation of the camera points without marking the targets, thereby achieving the technical effect of reducing the evaluation cost, thereby solving the problem of high evaluation cost caused by the large number of marked targets in the data processing method of the camera points in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0012] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0013] Figure 1 is a schematic diagram of a hardware environment of an optional camera point data processing method according to an embodiment of the present application;

[0014] Figure 2 It is a flowchart of an optional camera point data processing method according to an embodiment of the present application;

[0015] Figure 3 is a schematic diagram of an optional camera point data processing method according to an embodiment of the present application;

[0016] Figure 4 is a flow chart of another optional camera point data processing method according to an embodiment of the present application;

[0017] Figure 5 is a schematic diagram of another optional camera point data processing method according to an embodiment of the present application;

[0018] Figure 6 is a schematic diagram of another optional camera point data processing method according to an embodiment of the present application;

[0019] Figure 7 is a flowchart of another optional camera point data processing method according to an embodiment of the present application;

[0020] Figure 8 is a structural block diagram of an optional camera point data processing device according to an embodiment of the present application;

[0021] Fig. 9 It is a structural block diagram of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] According to one aspect of an embodiment of the present application, a method for processing camera point data is provided. Optionally, in this embodiment, the above-mentioned camera point data processing method can be applied to Figure 1 In the hardware environment including the shooting device 102 and the server 104 shown in FIG. Figure 1 As shown, the server 104 is connected to the shooting device 102 via a network, and can be used to provide services (such as application services, etc.) for the shooting device 102 or a client installed on the shooting device. A database can be set on the server or independently of the server to provide data storage services for the server 104.

[0025] The network may include but is not limited to at least one of the following: a wired network, a wireless network. The wired network may include but is not limited to at least one of the following: a wide area network, a metropolitan area network, a local area network, and the wireless network may include but is not limited to at least one of the following: WIFI (Wireless Fidelity), Bluetooth. The shooting device 102 may not be limited to a camera or the like.

[0026] The data processing method of the camera point in the embodiment of the present application can be executed by the server 104, or can be executed by the server 104 and the shooting device 102 together. Taking the case where the data processing method of the camera point in the embodiment is executed by the server 104 as an example, Figure 2 is a flow chart of an optional camera point data processing method according to an embodiment of the present application, such as Figure 2 As shown, the process of the method may include the following steps:

[0027] Step S202, parse the objects appearing in a group of snapshot data corresponding to a group of camera points to obtain a group of target objects and snapshot data corresponding to each target object in the group of target objects, wherein a group of camera points are points where cameras are installed in a preset area, a group of snapshot data includes the snapshot data of the camera at each camera point in the group of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data in which each target object appears in the group of snapshot data.

[0028] The data processing method of the camera point in this embodiment can be applied to the scenario of analyzing the construction quality of the camera points in a preset area. The preset area here can be a pre-set fixed area, an area of ​​a city, an area of ​​a district, or an area containing only a few streets, which is not limited in this embodiment. The preset area can contain multiple camera points. Here, the camera point can represent a point where a camera is installed. The camera point distribution diagram can be as follows: Figure 3 As shown, in Figure 3 In the figure, the cross star represents the position of the camera (ie, camera head).

[0029] Each city has deployed a certain scale of security system, including front-end cameras distributed in every corner of the city and corresponding back-end platform systems, such as Figure 4 As shown in the figure, the camera is responsible for real-time recording, capturing the target object, and other functions. The data generated by the camera is transmitted to the platform system, and the platform system completes the recording storage and data analysis. Under normal circumstances, due to the complexity and diversity of roads, a city often has thousands to tens of thousands of camera points. These cameras were built at different times. Some cameras have very low monitoring quality due to long-term operation. For example, the height of the camera is too high or too low, the camera angle is not easy to shoot the target, the camera is blocked by grown vegetation, etc. Therefore, it is necessary to timely evaluate the construction quality of the camera points in each area. The evaluation method is usually to use n marked targets to pass on the route where the camera is set up. Since the target, the route and the monitoring point are known, the construction quality of the camera on a route can be evaluated.

[0030] Although the above method is highly accurate, there are hundreds of roads in a city. If we evaluate each road one by one, the execution cost is too high and the evaluation time is lengthy. Usually, the evaluation results of a few routes are used to replace the evaluation results of the construction quality of camera points in the entire area or even the entire city to reduce the execution cost and evaluation time. However, this will reduce the accuracy of the results. At the same time, although this method reduces part of the evaluation cost, even if only a few routes are selected for evaluation, it still requires a certain amount of manpower and time costs, and the problem of high evaluation cost still exists.

[0031] At present, in the related technology, there is a method for evaluating the quality of the point according to the point density of the camera in the area. By obtaining multiple deployed camera points to construct a point triangulation network, the point triangulation network includes multiple point triangles, and the Thiessen polygons of the jurisdiction area corresponding to each deployed camera point are constructed based on the center of the circumscribed circle of the point triangle. The area of ​​the Thiessen polygons of the multiple jurisdiction areas is calculated, and a recommended site selection list is generated according to the area. The camera points to be deployed are included in the recommended site selection list. The larger the area of ​​the Thiessen polygon, the more it is preferred to be recommended as the camera point to be deployed. This method calculates the area of ​​the figure formed geometrically by the camera points. If the area is large and the monitoring points are sparse, then the new monitoring points are given priority. The priority is how to build new camera points, and it does not consider how to identify the cameras that need to be managed among the existing cameras, which is not conducive to the reuse of old cameras.

[0032] In addition, there is also a method in the related art to determine the quality of the camera point by analyzing the image of the monitoring point (camera point), parsing the pictures and videos of the monitoring point, and determining the quality level of the evaluation target type according to the semantic description information of the image parsed. The image semantic description information can not only evaluate the quality-related dimensions of the target image, but also comprehensively evaluate the installation height, angle, scene rationality and other point installation and deployment dimensions of the monitoring point, comprehensively and accurately feedback the quality of the monitoring point and warn of abnormal situations, thereby achieving more accurate monitoring point management. This method analyzes the image of the monitoring point, gives an image quality score, identifies the camera point that needs to be managed, and then manually verifies whether it needs to be managed. Although the point that needs to be managed is identified, it is required that the camera captures the image of the target before analysis can be performed. In real life, many cameras are of poor construction quality, not because the captured image quality is poor, but because the installation angle is unreasonable, or it is blocked by growing plants, etc., it is difficult to capture the image of the target, but this method does not involve such cameras.

[0033] In order to at least partially solve the above technical problems, in this embodiment, the overall camera point construction quality in the preset area can be determined based on the sum of the captured data of multiple target objects captured by the camera points in the preset area and the sum of the number of camera points involved in the actual movement trajectories of the multiple target objects corresponding to the captured data, thereby achieving a fast and low-cost evaluation of the construction instructions of camera points in a large area. Here, the target object can be a person, a car, or other objects that are easy to identify and distinguish. The above group of target objects can include only one target object or multiple target objects, which is not limited in this embodiment.

[0034] Since the number of target objects (people or cars) in a city can range from tens of thousands to millions, or even more, according to the "law of large numbers" in probability theory and mathematical statistics, although for a specific target object, the estimated value of the captured data of multiple target objects captured by camera points in a preset area may be biased, as long as the value of the target object is large enough, the error can be ignored and the overall pattern can be evaluated. When the selected time period is long enough, the number of target objects obtained in the preset area can meet the requirements of the law of large numbers.

[0035] When conducting a construction quality assessment of camera points, an object analysis can be performed on a set of snapshot data captured and uploaded by a set of camera points in a preset area within a preset time to determine the objects that appear in the preset area within the preset time. Here, a set of camera points can be points where cameras are installed in a preset area. A set of snapshot data can include the snapshot data of the camera at each camera point in a set of camera points within a preset time. The snapshot data can include the snapshot pictures taken by the camera, and can also include information such as the snapshot time and snapshot location corresponding to the snapshot pictures. The snapshot pictures can be individual pictures or videos composed of multiple frames of pictures. The preset time can be a continuous period of time, for example, a week, a month, etc.

[0036] In this embodiment, by analyzing the objects appearing in a group of snapshot data, a group of target objects and snapshot data corresponding to each target object in the group of target objects can be obtained. Here, the snapshot data corresponding to each target object can be the snapshot data appearing in each target object in the group of snapshot data.

[0037] Optionally, the above analysis method may be to perform object recognition and clustering on the same objects appearing in a group of captured data.

[0038] Step S204, fitting the movement trajectory of each target object according to the captured data corresponding to each target object, wherein the movement trajectory of each target object is used to indicate the predicted camera point that each target object passes through.

[0039] In this embodiment, the motion trajectory of each target object can be fitted based on the captured data corresponding to each target object. Here, the motion trajectory of each target object can be used to indicate the predicted camera point that each target object passes through.

[0040] Optionally, in order to avoid the situation where the fitted trajectory is actually impassable, when fitting the movement trajectory of each target object, each camera point in a set of camera points can be combined with the map data corresponding to the preset area, and the movement trajectory of each target object can be determined based on the communicative straight sections displayed on the map data and the position of the camera point corresponding to the snapshot data of each target object in the map data.

[0041] For example, Figure 3 As shown, after obtaining the public road network map, the camera locations can be marked on the map according to the latitude and longitude information of the camera points recorded when the new camera monitoring points are created in each area.

[0042] Optionally, taking into account the possibility that the target object may not be captured due to poor construction quality of individual camera points, resulting in no navigable straight section between the two camera points corresponding to the captured data of the target object, it is possible to consider taking the navigable winding section between the two camera points as part of the target object's movement trajectory.

[0043] Optionally, based on the determined movement trajectory of each target object and combined with the actual position of each camera point in a set of camera points, the camera points passed by each target object can be predicted.

[0044] Step S206, determines the total number of captured data contained in the captured data corresponding to each target object as the actual number of captured data of a group of camera points, and determines the total number of camera points contained in the movement trajectory of each target object as the predicted number of captured data of a group of camera points.

[0045] In this embodiment, the total number of snapshot data contained in the snapshot data corresponding to each target object can be determined as the actual number of snapshots for a group of camera points, and the total number of camera points contained in the movement trajectory of each target object can be determined as the predicted number of snapshots for a group of camera points.

[0046] The actual number of snapshots of the above-mentioned set of camera points may be the sum of the number of camera points that actually captured the target object in the snapshot data of each target object in the set of target objects. The predicted number of snapshots of the above-mentioned set of camera points may be the sum of the number of camera points that each target object in the set of target objects passes through theoretically determined based on the fitted motion trajectory.

[0047] Step S208, performing point anomaly prediction on a group of camera points according to the actual number of captured shots of a group of camera points and the predicted number of captured shots of a group of camera points, so as to predict whether there are abnormal camera points in the group of camera points.

[0048] The construction quality of the camera point can be determined by whether the camera can clearly capture the target and thus play a monitoring role. In this embodiment, a group of camera points can be predicted for point abnormality based on the actual number of snapshots of a group of camera points and the predicted number of snapshots of a group of camera points to predict whether there are abnormal camera points in the group of camera points. Here, point abnormality can mean that the camera point cannot capture the target and needs to be replaced and the installation position and angle need to be adjusted.

[0049] Optionally, in order to avoid inaccurate results of point abnormality prediction due to calculation errors and other reasons, the point abnormality prediction can be performed by combining the actual number of snapshots of a group of camera points of multiple target objects in a group of target objects and the predicted number of snapshots of a group of camera points. In addition, the above process can also be performed on a group of snapshot data corresponding to a group of camera points at multiple different times in a preset area.

[0050] Optionally, when the actual number of snapshots of a group of camera points of multiple target objects is inconsistent with the predicted number of snapshots of a group of camera points, it can be determined that there are abnormal camera points in the group of camera points. It can also be determined that there are abnormal camera points in the group of camera points when the actual number of snapshots of a group of camera points of multiple target objects is less than the predicted number of snapshots of a group of camera points.

[0051] Through the above steps, objects appearing in a set of snapshot data corresponding to a set of camera points are parsed to obtain a set of target objects and snapshot data corresponding to each target object in the set of target objects, wherein a set of camera points are points where cameras are installed in a preset area, a set of snapshot data includes snapshot data of the camera at each camera point in the set of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data of each target object appearing in the set of snapshot data; based on the snapshot data corresponding to each target object, a motion trajectory of each target object is fitted, wherein the motion trajectory of each target object is used to indicate the predicted motion trajectory of each target object. The invention discloses a method for detecting camera points passed by a target object; determining the total number of captured data contained in the captured data corresponding to each target object as the actual captured number of a group of camera points, and determining the total number of camera points contained in the movement trajectory of each target object as the predicted captured number of a group of camera points; performing point anomaly prediction on a group of camera points according to the actual captured number of a group of camera points and the predicted captured number of a group of camera points to predict whether there are abnormal camera points in a group of camera points, thereby solving the problem of high evaluation cost caused by a large number of marked targets in the data processing method of camera points in the related art and reducing the evaluation cost.

[0052] In an exemplary embodiment, objects appearing in a set of captured data corresponding to a set of camera points are parsed to obtain a set of target objects and captured data corresponding to each target object in the set of target objects, including:

[0053] S11, performing object recognition on a captured image in each captured data of a group of captured data to obtain a candidate object corresponding to each captured data;

[0054] S12, clustering the candidate objects corresponding to each captured data according to the object features of the candidate objects corresponding to each captured data, to obtain a group of target objects and the captured data corresponding to each target object.

[0055] Since each camera can capture multiple objects passing by the camera point within a preset time, correspondingly, there may be multiple objects in the captured data corresponding to each camera point in a group of camera points. When parsing the objects appearing in a group of captured data, object recognition can be performed on the captured pictures in each captured data. When the objects in each captured data are determined, the same objects are clustered into one class by clustering to obtain a group of target objects.

[0056] In this embodiment, object recognition can be performed on the captured images in each captured data of a group of captured data to obtain candidate objects corresponding to each captured data. Here, the candidate objects can be objects recognized in the captured images of each captured data, and the same candidate objects can be clustered to obtain the target object. The contents contained in the candidate objects can be similar to the description of the aforementioned objects, and this embodiment will not be repeated here.

[0057] Optionally, the object recognition may refer to parsing information such as feature values ​​and attributes of people and cars in the video / image according to a pre-set deep learning algorithm. The parsed data may be presented in a structured manner.

[0058] For example, taking face analysis as an example, the principles of other objects are similar. A typical structured information of a face can be shown in Table 1. According to the information in the captured data, the picture ID (Identity document), capture time and capture location of each picture can be determined. The corresponding feature value is determined according to the features of the object appearing in the picture.

[0059] Table 1

[0060]

[0061] For a city, many video / image streams can be analyzed at the same time. Even a small or medium-sized city can generate tens of millions of structured data in a day.

[0062] According to the object features of the candidate objects corresponding to each captured data, the candidate objects corresponding to each captured data can be clustered to obtain a group of target objects and the captured data corresponding to each target object. Here, each target object in the group of target objects can correspond to multiple candidate objects.

[0063] Optionally, when clustering the candidate objects corresponding to each captured data, the candidate objects can be clustered by a preset clustering algorithm. Similar candidate objects are determined by calculating the similarity of the feature values ​​of each candidate object, and the similar candidate objects are used as a target object. Here, the similarity calculation formula for the feature value can be a cosine similarity calculation formula or an L1 (Manhattan distance) calculation formula.

[0064] For example, taking the cosine similarity calculation formula as an example, the cosine similarity calculation formula can be shown as formula (1), and the value of cosine similarity is [-1, 1]. The larger the value, the more similar it is.

[0065]

[0066] Among them, A i,B i can represent the components of the eigenvectors A and B respectively.

[0067] Optionally, based on the clustering results, a file can be created for each target object, recording information such as the target object's picture, capture time and capture location, so as to facilitate the restoration of the target object's movement trajectory based on the capture time and capture location information (i.e., space-time information) in the file.

[0068] Optionally, when the number of candidate images in a class in the clustering result is small, in order to avoid trajectory fitting failure or low reference value of the fitted trajectory due to too small number of candidate images, the candidate images of this class may be directly discarded.

[0069] Through this embodiment, by clustering the target objects based on the identified object features in the captured images, the clustering accuracy and efficiency of the target objects can be improved.

[0070] In an exemplary embodiment, fitting the motion trajectory of each target object according to the captured data corresponding to each target object includes:

[0071] S21, taking each target object as the current target object and performing the following moving trajectory fitting operation to obtain the moving trajectory of each target object, wherein when performing the following moving trajectory fitting operation, the captured data corresponding to the current target object is a set of current captured data:

[0072] Determine a camera point sequence that the current target object passes through according to the spatiotemporal information of each current captured data in a set of current captured data, wherein the spatiotemporal information of each current captured data includes a camera point corresponding to each current captured data and a capture time corresponding to each current captured data;

[0073] In the area map corresponding to the preset area, when there is a direct road between any adjacent camera points in the camera point sequence, the travel trajectory represented by the camera point sequence is determined as the action trajectory of the current target object;

[0074] In a regional map corresponding to a preset area, when there is no direct road between adjacent first camera points and second camera points in a camera point sequence, a traversable route between the first camera point and the second camera point in the regional map is determined to obtain a predicted traversable route; the camera points on the predicted traversable route are added between the first camera point and the second camera point in the camera point sequence to obtain an updated camera point sequence; and the travel trajectory represented by the updated camera point sequence is determined as the action trajectory of each target object.

[0075] After determining a group of target objects and the snapshot data corresponding to each target object, the movement trajectory of each target object can be fitted according to the spatiotemporal information of the snapshot data corresponding to each target object. Here, the spatiotemporal information of the snapshot data corresponding to each target object can be the camera point (i.e., the snapshot position) and the snapshot time recorded in the snapshot data corresponding to the snapshot data.

[0076] In this embodiment, each target object can be used as the current target object to perform a moving trajectory fitting operation to obtain the movement trajectory of each target object. Correspondingly, the captured data corresponding to the current target object is a set of current captured data. The spatiotemporal information of each current captured data may include the camera point corresponding to each current captured data and the capture time corresponding to each current captured data.

[0077] The moving trajectory fitting operation may be to determine the camera point sequence that the current target object passes through according to the spatiotemporal information of each current captured data in a set of current captured data, and determine the corresponding action trajectory according to the camera point sequence. Here, the camera point sequence may be the sequence of camera points determined according to the sequence of capture time.

[0078] In the regional map corresponding to the preset area, if there is a direct road between any adjacent camera points in the camera point sequence, the travel trajectory represented by the camera point sequence can be determined as the action trajectory of the current target object. Here, the regional map can be a map that matches the actual road conditions of the preset area, and the camera point information can be displayed in the regional map.

[0079] Considering that some cameras in the movement trajectory of the front target object may not capture the target object due to poor construction quality, etc., resulting in no direct road between two adjacent camera points in the camera point sequence. In this embodiment, in the regional map corresponding to the preset area, if there is no direct road between the adjacent first camera point and the second camera point in the camera point sequence, a passable route between the first camera point and the second camera point in the regional map can be determined to obtain a predicted passable route.

[0080] Optionally, when there are multiple passable routes between the first camera point and the second camera point, the predicted passable route may be the shortest passable route between the first camera point and the second camera point, or may be a passable route with the best traffic conditions among the passable routes between the first camera point and the second camera point, or may be any one of the passable routes between the first camera point and the second camera point, which is not limited in this embodiment.

[0081] After determining the predicted route, the camera points on the predicted route can be added between the first camera point and the second camera point in the camera point sequence to obtain an updated camera point sequence, and the travel trajectory represented by the updated camera point sequence can be determined as the action trajectory of each target object.

[0082] For example, taking the target object as target p1, after creating a file for target p1 according to the clustering result, the number of points captured in the file is counted, such as Figure 5 As shown in the figure, there are snapshots at points A, B, D, F, and G. Combined with the road conditions in the map data, it can be determined that there is no direct road between BD and DF. The reason may be that the camera points on other roads between BD or DF are not well constructed and the target is not captured. According to the road conditions shown in the map data, it can be determined that there are multiple accessible routes between BD and DF. A accessible route can be randomly selected as the trajectory of the target p1, or it can be selected based on other criteria.

[0083] Through this embodiment, the movement trajectory of the target object is determined according to the camera point position displayed by the captured data of the target object in combination with the road conditions in the map, so that the fitting accuracy of the movement trajectory can be improved.

[0084] In an exemplary embodiment, in a regional map corresponding to a preset area, when there is no direct road between adjacent first camera points and second camera points in a camera point sequence, determining a traversable route between the first camera point and the second camera point in the regional map to obtain a predicted traversable route includes:

[0085] S31, in the area map corresponding to the preset area, when there is no direct road between the adjacent first camera point and the second camera point in the camera point sequence, and there are multiple passable routes between the first camera point and the second camera point, predict the shortest passable route between the first camera point and the second camera point in the area map to obtain the predicted passable route.

[0086] In the area map corresponding to the preset area, when there is no direct road between the adjacent first camera point and the second camera point in the camera point sequence, and there are multiple passable routes between the first camera point and the second camera point, when determining the predicted passable route, the shortest passable route among the multiple passable routes between the first camera point and the second camera point can be selected as the determined predicted passable route.

[0087] Optionally, the shortest traversable route can be selected by abstracting multiple traversable routes between the first camera point and the second camera point and converting them into a problem of solving the shortest path of a graph. In this embodiment, the algorithm for solving the path can be a Dijkstra (also known as the shortest path search method) algorithm, a Floyd (also known as the insertion point method) algorithm, or a SPFA (Shortest Path Faster Algorithm) algorithm or other algorithms.

[0088] For example, taking Dijkstra's algorithm as an example, Figure 6 As shown, the edge has a weight, and the weight is the physical distance between the two camera points. The basic process of the Dijkstra algorithm can be, first divide the vertex set V into two groups, namely S: the set of vertices for which the shortest path has been found (initially only contains the source point v0); VS = T: the set of vertices for which the shortest path has not yet been determined. Then add the vertices in T to S in ascending order, ensuring that: the length from the source point v0 to the other vertices in S is not greater than the shortest path length from v0 to any vertex in T, and each vertex corresponds to a distance value (vertex in S: the length from v0 to this vertex, vertex in T: the shortest path length from v0 to this vertex that only includes vertices in S as intermediate vertices).

[0089] like Figure 5 As shown in the figure, the ABCDEFG route is fitted based on the ABDFG route. It is the shortest road that the target is most likely to travel. Based on this fitted route, the maximum value of the camera point position of p1 passing through the ABCDEFG route can be counted.

[0090] Through this embodiment, the shortest travel path is selected as the predicted travel route when there is no directly connected road between adjacent camera points, which can improve the matching degree between the predicted travel route and the actual route.

[0091] In an exemplary embodiment, according to the actual number of snapshots of a group of camera points and the predicted number of snapshots of a group of camera points, point abnormality prediction is performed on a group of camera points, including:

[0092] S41, determining a ratio between a predicted number of captured images of a group of camera points and an actual number of captured images of a group of camera points to obtain a first reference ratio;

[0093] S42, when the first reference ratio is greater than or equal to the first preset ratio, determining that there is no abnormal camera point in the group of camera points;

[0094] S43: When the first reference ratio is less than the first preset ratio, determining that there is an abnormal camera point in a group of camera points.

[0095] When predicting point anomalies for a group of camera points based on the actual number of snapshots of a group of camera points and the predicted number of snapshots of a group of camera points, the ratio between the predicted number of snapshots of a group of camera points and the actual number of snapshots of a group of camera points can be calculated, and the quantitative processing of evaluating the construction quality of camera points in the preset area can be achieved by converting the construction quality of camera points in the preset area into mathematically measurable numbers.

[0096] In this embodiment, a first preset ratio may be pre-set to be used for comparison with the ratio between the predicted number of snapshots of a group of camera points and the actual number of snapshots of a group of camera points, thereby determining the construction quality of the camera points in the preset area. The ratio between the determined predicted number of snapshots of a group of camera points and the actual number of snapshots of a group of camera points may be recorded as the first reference ratio.

[0097] The first reference ratio is compared with the first preset ratio. If the first reference ratio is greater than or equal to the first preset ratio, it can be determined that there is no abnormal camera point in the group of camera points. Correspondingly, if the first reference ratio is less than the first preset ratio, it can be determined that there is an abnormal camera point in the group of camera points.

[0098] For example, within a limited period of time, count p1, p2, ... p k The number of camera points that the target object theoretically passes through and the camera point data that is actually captured in the archive are recorded as n1, n2, ... n respectively. k and m1,m2,…m k The score of the camera point construction quality in a preset area can be expressed as formula (2). The better the camera point construction quality in the preset area, the closer the score is to 1.

[0099]

[0100] Through this embodiment, the construction quality of camera points in a preset area is determined by the ratio between the predicted number of snapshots of a group of camera points and the actual number of snapshots of a group of camera points, thereby realizing the transformation of the construction quality of camera points in a preset area into mathematically measurable numbers, which can improve the efficiency of evaluating the construction quality of camera points in a preset area.

[0101] In an exemplary embodiment, after performing point abnormality prediction for a group of camera points according to the actual number of snapshots of a group of camera points and the predicted number of snapshots of a group of camera points, the method further includes:

[0102] S51, when it is predicted that there is an abnormal camera point in a group of camera points, determine the number of captured data belonging to each camera point in the captured data corresponding to each target object, obtain the actual number of captured data of each camera point, and determine the total number of each camera point appearing in the action trajectory of each target object, and obtain the predicted number of captured data of each camera point;

[0103] S52, performing point abnormality prediction for each camera point according to the actual number of captured shots of each camera point and the predicted number of captured shots of each camera point, so as to predict whether each camera point is an abnormal camera point.

[0104] After determining that there are abnormal camera points in a set of camera points in a preset area, the camera points with abnormalities may be further evaluated and determined. In this embodiment, the construction quality of each camera point in a set of camera points may be evaluated in a manner similar to the manner of evaluating the construction quality of a set of camera points in a preset area.

[0105] When an abnormal camera point is predicted in a group of camera points, the number of snapshot data belonging to each camera point in the snapshot data corresponding to each target object can be determined to obtain the actual number of snapshots of each camera point, and the total number of each camera point appearing in the action trajectory of each target object can be determined to obtain the predicted number of snapshots of each camera point. Here, the actual number of snapshots of each camera point corresponds to the snapshot data, and the predicted number of snapshots of each camera point corresponds to the action trajectory of a group of target objects.

[0106] According to the actual number of snapshots of each camera point and the predicted number of snapshots of each camera point, point anomaly prediction can be performed for each camera point to predict whether each camera point is an abnormal camera point.

[0107] Optionally, when the actual number of snapshots at each camera point is inconsistent with the predicted number of snapshots at each camera point, or the difference is large, the camera point can be determined to be an abnormal camera point.

[0108] Optionally, other methods may be used to complete the evaluation of the construction quality of each camera point in a group of camera points. For example, the quality of the captured images uploaded by each camera point in a group of camera points within a preset time is analyzed, and the camera point with multiple captured images of poor quality is determined as an abnormal camera point.

[0109] Through this embodiment, abnormal cameras are determined based on the predicted number of snapshots of each camera point and the actual number of snapshots of each camera point. Since the processed data for evaluating the construction quality of a group of camera points in a preset area can be directly used, the resource utilization rate can be improved.

[0110] In an exemplary embodiment, according to the actual number of snapshots of each camera point and the predicted number of snapshots of each camera point, point abnormality prediction is performed for each camera point, including:

[0111] S61, taking each camera point as the current camera point and performing the following point abnormality prediction operation:

[0112] Determine the ratio between the predicted number of snapshots at the current camera point and the actual number of snapshots at the current camera point to obtain a second reference ratio;

[0113] When the second reference ratio is greater than or equal to the second preset ratio, determining that the current camera position is not an abnormal camera position;

[0114] When the second reference ratio is less than the second preset ratio, it is determined that the current camera position is an abnormal camera position.

[0115] When performing point anomaly prediction for each camera point based on the actual number of snapshots of each camera point and the predicted number of snapshots of each camera point, each camera point can be used as the current camera point to perform the point anomaly prediction operation.

[0116] The above-mentioned point abnormality prediction operation may include: determining the ratio between the predicted number of snapshots of the current camera point and the actual number of snapshots of the current camera point, obtaining a second reference ratio, and comparing the second reference ratio with a second preset ratio. Here, the second preset ratio may be a preset value, which may be the same as or different from the value of the first preset ratio, and this embodiment does not limit this.

[0117] In this embodiment, when the second reference ratio is greater than or equal to the second preset ratio, it can be determined that the current camera position is not an abnormal camera position. When the second reference ratio is less than the second preset ratio, it can be determined that the current camera position is an abnormal camera position.

[0118] Through this embodiment, according to the ratio between the predicted number of snapshots of the current camera point and the actual number of snapshots of the current camera point, a quantitative evaluation of the construction quality of a single camera point can be achieved.

[0119] The following is an explanation of the data processing method of the camera point in the embodiment of the present application in conjunction with an optional example. In this optional example, the target object is a target.

[0120] In the related art, the evaluation of the construction quality of camera points in a large area usually involves selecting several routes for setting up cameras, using n marked targets to pass through the selected routes, and determining the construction quality of the camera points on the corresponding routes based on the obtained monitoring results, and then determining the construction quality of the camera points in the entire area. This method requires a large investment and takes a long time. In this regard, this optional example provides a method for evaluating the construction quality of camera points. By fitting the routes that multiple targets are most likely to pass through, and counting the cameras on the corresponding routes, the overall camera point construction quality score of the area is calculated. Without investing a lot of manpower costs, the construction quality of camera points in a large area can be quickly and accurately evaluated.

[0121] like Figure 7 As shown, the process of the camera point data processing method in this optional example may include the following steps:

[0122] Step S702, obtain a road network map and mark the camera location on the map.

[0123] Step S704: clustering the objects appearing in the video / image using a deep learning algorithm.

[0124] Step S706, counting the number of camera points that the target theoretically passes through and the number of camera points that have captured the image in the archive.

[0125] Step S708, calculating the overall camera point construction quality score.

[0126] Through this optional example, the quality of camera point construction in an area can be determined through the number of camera points that multiple targets theoretically pass through and the number of captured camera points in the archives, without the need for human intervention. While reducing the evaluation cost, the quality of camera point construction is turned into a mathematically measurable number, thus achieving quantitative processing of camera point construction.

[0127] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0128] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM (Read-Only Memory) / RAM (Random Access Memory), a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of each embodiment of the present application.

[0129] According to another aspect of an embodiment of the present application, a camera point data processing device for implementing the above-mentioned camera point data processing method is also provided. Figure 8 is a structural block diagram of an optional camera point data processing device according to an embodiment of the present application, such as Figure 8 As shown, the device may include:

[0130] The parsing unit 802 is used to parse the objects appearing in a set of captured data corresponding to a set of camera points to obtain a set of target objects and captured data corresponding to each target object in the set of target objects, wherein the set of camera points are points where cameras are installed in a preset area, the set of captured data includes the captured data of the camera at each camera point in the set of camera points within a preset time, and the captured data corresponding to each target object is the captured data of each target object appearing in the set of captured data;

[0131] The fitting unit 804 is connected to the parsing unit 802 and is used to fit the motion trajectory of each target object according to the captured data corresponding to each target object, wherein the motion trajectory of each target object is used to indicate the predicted camera point that each target object passes through;

[0132] The first determination unit 806 is connected to the fitting unit 804, and is used to determine the total number of captured data contained in the captured data corresponding to each target object as the actual number of captured data of a group of camera points, and to determine the total number of camera points contained in the movement trajectory of each target object as the predicted number of captured data of a group of camera points.

[0133] The first prediction unit 808 is connected to the first determination unit 806, and is used to perform point anomaly prediction on a group of camera points based on the actual number of snapshots of a group of camera points and the predicted number of snapshots of a group of camera points, so as to predict whether there are abnormal camera points in the group of camera points.

[0134] It should be noted that the parsing unit 802 in this embodiment can be used to execute the above step S202, the fitting unit 804 in this embodiment can be used to execute the above step S204, the first determination unit 806 in this embodiment can be used to execute the above step S206, and the first prediction unit 808 can be used to execute the above step S208.

[0135] Through the above module, objects appearing in a set of snapshot data corresponding to a set of camera points are parsed to obtain a set of target objects and snapshot data corresponding to each target object in the set of target objects, wherein a set of camera points are points where cameras are installed in a preset area, a set of snapshot data includes snapshot data of the camera at each camera point in the set of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data of each target object appearing in the set of snapshot data; based on the snapshot data corresponding to each target object, the movement trajectory of each target object is fitted, wherein the movement trajectory of each target object is used to indicate the predicted movement trajectory of each target object. The camera points passed by the target object; the total number of snapshot data contained in the snapshot data corresponding to each target object is determined as the actual snapshot number of a group of camera points, and the total number of camera points contained in the movement trajectory of each target object is determined as the predicted snapshot number of a group of camera points; according to the actual snapshot number of a group of camera points and the predicted snapshot number of a group of camera points, a point anomaly prediction is performed on a group of camera points to predict whether there are abnormal camera points in a group of camera points, which solves the problem of high evaluation cost caused by a large number of marked targets in the data processing method of camera points in the related art, thereby reducing the evaluation cost.

[0136] In an exemplary embodiment, the parsing unit includes:

[0137] A recognition module, used for performing object recognition on a captured image in each captured data of a group of captured data, and obtaining a candidate object corresponding to each captured data;

[0138] The clustering module is used to cluster the candidate objects corresponding to each captured data according to the object features of the candidate objects corresponding to each captured data, so as to obtain a group of target objects and the captured data corresponding to each target object.

[0139] In an exemplary embodiment, the fitting unit comprises:

[0140] The first execution module is used to perform the following moving track fitting operation on each target object as the current target object to obtain the moving track of each target object, wherein when performing the following moving track fitting operation, the captured data corresponding to the current target object is a set of current captured data:

[0141] Determine a camera point sequence that the current target object passes through according to the spatiotemporal information of each current captured data in a set of current captured data, wherein the spatiotemporal information of each current captured data includes a camera point corresponding to each current captured data and a capture time corresponding to each current captured data;

[0142] In the area map corresponding to the preset area, when there is a direct road between any adjacent camera points in the camera point sequence, the travel trajectory represented by the camera point sequence is determined as the action trajectory of the current target object;

[0143] In a regional map corresponding to a preset area, when there is no direct road between adjacent first camera points and second camera points in a camera point sequence, a traversable route between the first camera point and the second camera point in the regional map is determined to obtain a predicted traversable route; the camera points on the predicted traversable route are added between the first camera point and the second camera point in the camera point sequence to obtain an updated camera point sequence; and the travel trajectory represented by the updated camera point sequence is determined as the action trajectory of each target object.

[0144] In an exemplary embodiment, the first execution module includes:

[0145] The prediction submodule is used to predict the shortest passable route between the first camera point and the second camera point in the regional map corresponding to the preset area, when there is no direct road between the adjacent first camera point and the second camera point in the camera point sequence, and there are multiple passable routes between the first camera point and the second camera point, so as to obtain the predicted passable route.

[0146] In an exemplary embodiment, the first prediction unit includes:

[0147] A first determination module is used to determine a ratio between a predicted number of snapshots of a group of camera points and an actual number of snapshots of a group of camera points to obtain a first reference ratio;

[0148] A second determination module is used to determine that there is no abnormal camera point in a group of camera points when the first reference ratio is greater than or equal to the first preset ratio;

[0149] The third determination module is used to determine that there is an abnormal camera point in a group of camera points when the first reference ratio is less than the first preset ratio.

[0150] In an exemplary embodiment, the above device further comprises:

[0151] A second determination unit is used for, after performing point abnormality prediction on a group of camera points according to an actual number of snapshots of a group of camera points and a predicted number of snapshots of a group of camera points, determining the number of snapshot data belonging to each camera point in the snapshot data corresponding to each target object, when it is predicted that there is an abnormal camera point in the group of camera points, to obtain the actual number of snapshots of each camera point, and determining the total number of each camera point appearing in the movement trajectory of each target object, to obtain the predicted number of snapshots of each camera point;

[0152] The second prediction unit is used to perform point abnormality prediction for each camera point according to the actual number of snapshots of each camera point and the predicted number of snapshots of each camera point, so as to predict whether each camera point is an abnormal camera point.

[0153] In an exemplary embodiment, the second prediction unit includes:

[0154] The second execution module is used to perform the following point abnormality prediction operations on each camera point as the current camera point:

[0155] Determine the ratio between the predicted number of snapshots at the current camera point and the actual number of snapshots at the current camera point to obtain a second reference ratio;

[0156] When the second reference ratio is greater than or equal to the second preset ratio, determining that the current camera position is not an abnormal camera position;

[0157] When the second reference ratio is less than the second preset ratio, it is determined that the current camera position is an abnormal camera position.

[0158] It should be noted that the examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the contents disclosed in the above embodiments. It should be noted that the above modules as part of the device can be run in Figure 1 In the hardware environment shown, it can be implemented by software or by hardware, wherein the hardware environment includes a network environment.

[0159] According to another aspect of the embodiments of the present application, a storage medium is further provided. Optionally, in this embodiment, the storage medium can be used to execute the program code of any of the camera point data processing methods in the embodiments of the present application.

[0160] Optionally, in this embodiment, the storage medium may be located on at least one network device among a plurality of network devices in the network shown in the above embodiment.

[0161] Optionally, in this embodiment, the storage medium is configured to store program codes for executing the following steps:

[0162] S1, parsing objects appearing in a set of captured data corresponding to a set of camera points to obtain a set of target objects and captured data corresponding to each target object in the set of target objects, wherein a set of camera points are points where cameras are installed in a preset area, a set of captured data includes captured data of a camera at each camera point in the set of camera points within a preset time, and the captured data corresponding to each target object is captured data of each target object appearing in the set of captured data;

[0163] S2, fitting a motion trajectory of each target object according to the captured data corresponding to each target object, wherein the motion trajectory of each target object is used to indicate the predicted camera point that each target object passes through;

[0164] S3, determining the total number of captured data included in the captured data corresponding to each target object as the actual number of captured images of a group of camera points, and determining the total number of camera points included in the movement trajectory of each target object as the predicted number of captured images of a group of camera points;

[0165] S4, performing point anomaly prediction on a group of camera points according to an actual number of captured shots of a group of camera points and a predicted number of captured shots of a group of camera points, so as to predict whether there are abnormal camera points in the group of camera points.

[0166] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, which will not be described in detail in this embodiment.

[0167] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a ROM, a RAM, a mobile hard disk, a magnetic disk, or an optical disk.

[0168] According to another aspect of an embodiment of the present application, an electronic device for implementing the above-mentioned camera point data processing method is also provided. The electronic device may be a server, a terminal, or a combination thereof.

[0169] Fig. 9 is a structural block diagram of an optional electronic device according to an embodiment of the present application, such as Fig. 9As shown, it includes a processor 902, a communication interface 904, a memory 906 and a communication bus 908, wherein the processor 902, the communication interface 904 and the memory 906 communicate with each other through the communication bus 908, wherein,

[0170] Memory 906, used for storing computer programs;

[0171] The processor 902 is used to implement the following steps when executing the computer program stored in the memory 906:

[0172] S1, parsing objects appearing in a set of captured data corresponding to a set of camera points to obtain a set of target objects and captured data corresponding to each target object in the set of target objects, wherein a set of camera points are points where cameras are installed in a preset area, a set of captured data includes captured data of a camera at each camera point in the set of camera points within a preset time, and the captured data corresponding to each target object is captured data of each target object appearing in the set of captured data;

[0173] S2, fitting a motion trajectory of each target object according to the captured data corresponding to each target object, wherein the motion trajectory of each target object is used to indicate the predicted camera point that each target object passes through;

[0174] S3, determining the total number of captured data included in the captured data corresponding to each target object as the actual number of captured images of a group of camera points, and determining the total number of camera points included in the movement trajectory of each target object as the predicted number of captured images of a group of camera points;

[0175] S4, performing point anomaly prediction on a group of camera points according to an actual number of captured shots of a group of camera points and a predicted number of captured shots of a group of camera points, so as to predict whether there are abnormal camera points in the group of camera points.

[0176] Optionally, in this embodiment, the communication bus may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig. 9 The communication interface is used for communication between the electronic device and other devices.

[0177] The memory may include RAM, or may include non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0178] As an example, the memory 906 may include, but is not limited to, the parsing unit 802, the fitting unit 804, the first determination unit 806, and the first prediction unit 808 in the data processing device for the camera point. In addition, other module units in the data processing device for the camera point may also be included but are not limited to, which will not be repeated in this example.

[0179] The above-mentioned processor can be a general-purpose processor, which can include but not be limited to: CPU (Central Processing Unit), NP (Network Processor), etc.; it can also be DSP (Digital Signal Processing), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0180] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment will not be described in detail here.

[0181] It can be understood by those skilled in the art that Fig. 9 The structure shown is for illustration only. The device for implementing the above-mentioned camera point data processing method may be a terminal device, which may be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, MID), a PAD, and other terminal devices. Fig. 9 The structure of the electronic device is not limited. Fig. 9 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Fig. 9 Different configurations shown.

[0182] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which can include: a flash drive, ROM, RAM, a magnetic disk or an optical disk, etc.

[0183] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0184] If the integrated units in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application.

[0185] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0186] In the several embodiments provided in the present application, it should be understood that the disclosed client can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0187] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution provided in this embodiment.

[0188] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0189] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for processing camera point data, characterized in that: include: Parsing objects appearing in a set of snapshot data corresponding to a set of camera points to obtain a set of target objects and snapshot data corresponding to each target object in the set of target objects, wherein the set of camera points are points where cameras are installed in a preset area, the set of snapshot data includes snapshot data of a camera at each camera point in the set of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data in which each target object appears in the set of snapshot data; Fitting a movement trajectory of each target object according to the captured data corresponding to each target object, wherein the movement trajectory of each target object is used to indicate a predicted camera point passed by each target object; Determine the total number of captured data contained in the captured data corresponding to each target object as the actual number of captured data of the group of camera points, and determine the total number of camera points contained in the movement trajectory of each target object as the predicted number of captured data of the group of camera points; According to the actual number of snapshots of the group of camera points and the predicted number of snapshots of the group of camera points, point anomaly prediction is performed on the group of camera points to predict whether there are abnormal camera points in the group of camera points.

2. The method according to claim 1, characterized in that: The object appearing in a group of captured data corresponding to a group of camera points is parsed to obtain a group of target objects and captured data corresponding to each target object in the group of target objects, including: Performing object recognition on the captured images in each captured data of the set of captured data to obtain candidate objects corresponding to each captured data; The candidate objects corresponding to each captured data are clustered according to the object features of the candidate objects corresponding to each captured data to obtain the group of target objects and the captured data corresponding to each target object.

3. The method according to claim 1, characterized in that The step of fitting the movement trajectory of each target object according to the captured data corresponding to each target object includes: The following moving track fitting operation is performed on each target object as the current target object to obtain the moving track of each target object, wherein when the following moving track fitting operation is performed, the captured data corresponding to the current target object is a set of current captured data: Determine the camera point sequence passed by the current target object according to the spatiotemporal information of each current captured data in the set of current captured data, wherein the spatiotemporal information of each current captured data includes the camera point corresponding to each current captured data and the capturing time corresponding to each current captured data; In the area map corresponding to the preset area, when there is a direct road between any adjacent camera points in the camera point sequence, the travel trajectory represented by the camera point sequence is determined as the movement trajectory of the current target object; In the area map corresponding to the preset area, when there is no direct road between the adjacent first camera point and the second camera point in the camera point sequence, a traversable route between the first camera point and the second camera point in the area map is determined to obtain a predicted traversable route; the camera points on the predicted traversable route are added between the first camera point and the second camera point in the camera point sequence to obtain an updated camera point sequence; and the travel trajectory represented by the updated camera point sequence is determined as the action trajectory of each target object.

4. The method according to claim 3, characterized in that In the case where there is no direct road between adjacent first camera points and second camera points in the camera point sequence in the regional map corresponding to the preset area, determining a traversable route between the first camera point and the second camera point in the regional map to obtain a predicted traversable route includes: In the area map corresponding to the preset area, when there is no direct road between the adjacent first camera point and the second camera point in the camera point sequence, and there are multiple traversable routes between the first camera point and the second camera point, the shortest traversable route between the first camera point and the second camera point in the area map is predicted to obtain a predicted traversable route.

5. The method according to claim 1, characterized in that The step of performing point abnormality prediction on the group of camera points according to the actual number of captured shots of the group of camera points and the predicted number of captured shots of the group of camera points includes: Determine a ratio between a predicted number of snapshots of the set of camera points and an actual number of snapshots of the set of camera points to obtain a first reference ratio; When the first reference ratio is greater than or equal to a first preset ratio, determining that there is no abnormal camera point in the group of camera points; When the first reference ratio is less than the first preset ratio, it is determined that there are abnormal camera points in the group of camera points.

6. The method according to any one of claims 1 to 5, characterized in that After performing point abnormality prediction on the group of camera points according to the actual number of snapshots of the group of camera points and the predicted number of snapshots of the group of camera points, the method further includes: In the case where it is predicted that there is an abnormal camera point in the group of camera points, the number of captured data belonging to each camera point in the captured data corresponding to each target object is determined to obtain the actual number of captured data of each camera point, and the total number of each camera point appearing in the movement trajectory of each target object is determined to obtain the predicted number of captured data of each camera point; According to the actual number of snapshots of each camera point and the predicted number of snapshots of each camera point, point abnormality prediction is performed on each camera point to predict whether each camera point is an abnormal camera point.

7. The method according to claim 6, characterized in that The performing point abnormality prediction for each camera point according to the actual number of captured shots of each camera point and the predicted number of captured shots of each camera point includes: Take each camera point as the current camera point and perform the following point abnormality prediction operation: Determine a ratio between the predicted number of snapshots of the current camera point and the actual number of snapshots of the current camera point to obtain a second reference ratio; When the second reference ratio is greater than or equal to a second preset ratio, determining that the current camera position is not an abnormal camera position; When the second reference ratio is less than a second preset ratio, it is determined that the current camera position is an abnormal camera position.

8. A camera point data processing device, characterized in that: include: A parsing unit, used for parsing objects appearing in a set of snapshot data corresponding to a set of camera points, to obtain a set of target objects and snapshot data corresponding to each target object in the set of target objects, wherein the set of camera points are points where cameras are installed in a preset area, the set of snapshot data includes snapshot data of a camera at each camera point in the set of camera points within a preset time, and the snapshot data corresponding to each target object is the snapshot data in which each target object appears in the set of snapshot data; A fitting unit, used to fit the movement trajectory of each target object according to the captured data corresponding to each target object, wherein the movement trajectory of each target object is used to indicate the predicted camera point passed by each target object; A first determining unit is used to determine the total number of captured data included in the captured data corresponding to each target object as the actual number of captured images of the group of camera points, and to determine the total number of camera points included in the movement trajectory of each target object as the predicted number of captured images of the group of camera points; The first prediction unit is used to perform point anomaly prediction on the group of camera points according to the actual number of snapshots of the group of camera points and the predicted number of snapshots of the group of camera points, so as to predict whether there are abnormal camera points in the group of camera points.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein the program executes the method according to any one of claims 1 to 7 when executed.

10. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to any one of claims 1 to 7 through the computer program.

Images