A monitoring object processing method and device based on a cache mechanism

Abstract

The application relates to a monitoring object processing method and device based on a cache mechanism, which comprises the following steps: receiving a feature corresponding to a monitoring object, and inputting the feature into a region cache for storage; receiving a target feature corresponding to a current monitoring object output by a computing device; matching the target feature with the stored feature; if the matching fails, establishing a monitoring file corresponding to the current monitoring object in a feature library; and through a main control program, a region cache for storing feature comparison results is additionally arranged locally, so that the received target feature is compared with the stored feature in the region cache to determine whether the current monitoring object corresponding to the target feature is a new image feature. Through the matching processing in the region cache, the monitoring object missing detection caused by the asynchronous processing in the computing device and the failure of the feature library to update in time can be prevented.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a monitoring object processing method and device based on a cache mechanism and a computer device. BACKGROUND

[0002] Currently, monitoring devices for public areas can be seen everywhere, and the number of monitoring objects obtained thereby is close to a mass. Subsequent processing of continuous identification of specific objects from numerous monitoring objects is severely dependent on software processing time sequence archiving operations.

[0003] In the processing time sequence archiving operation, a concurrent processing mode is often used, and multiple monitoring images for the same target object are obtained at the same time. If the first monitoring image sequentially undergoes characteristic comparison, record archiving and other operations to complete the processing process, the second monitoring image is still in the characteristic comparison stage. At this time, since the monitoring object corresponding to the first monitoring image has not been added to the feature library, the second monitoring image cannot find the corresponding monitoring object in the feature library, and therefore the first and second monitoring images are each regarded as a new monitoring object and an archive is established.

[0004] According to the above processing mechanism, once the number of concurrent processing monitoring images is too large, a large number of monitoring objects of the same nature may be established as independent monitoring objects and archives due to a short processing time difference. Such a processing method causes the object recognition mechanism based on the monitoring object to fail, which seriously affects the normal use of the monitoring device. SUMMARY

[0005] Therefore, it is necessary to provide a monitoring object processing method and device based on a cache mechanism and a computer device to reduce the probability of failing to identify repeated monitoring objects caused by simply relying on the feature library recognition mechanism through an additional secondary matching operation.

[0006] In a first aspect, the present application provides a monitoring object processing method based on a cache mechanism, which comprises:

[0007] receiving a feature corresponding to a monitoring object, and inputting the feature to a region cache for storage, wherein the feature is output by a computing device after comparison with a feature library;

[0008] receiving a target feature corresponding to a current monitoring object output by the computing device;

[0009] matching the target feature with the stored feature;

[0010] if the matching fails, establishing a monitoring archive corresponding to the current monitoring object in the feature library.

[0011] In one of the embodiments, the receiving the feature of the corresponding monitoring object, inputting the feature to the region cache for storage, comprises:

[0012] Obtaining a target data structure, and constructing a region cache according to the target region structure;

[0013] Receiving the feature of the corresponding monitoring object after the computing device compares the feature library;

[0014] Storing the feature in the region cache in chronological order of generation.

[0015] In one of the embodiments, the obtaining a target data structure, and constructing a region cache according to the target region structure, comprises:

[0016] Reading position information of a monitoring device used for shooting the monitoring object;

[0017] Obtaining attribute information of the monitoring object;

[0018] Constructing a target data structure based on a data structure of the position information and a data structure of the attribute information;

[0019] Constructing a region cache according to the target data structure.

[0020] In one of the embodiments, the matching the target feature with the stored feature, comprises:

[0021] Selecting the feature and the target feature for operation, and selecting a to-be-processed feature corresponding to the target feature based on the operation result;

[0022] Performing time-dimension filtering processing on the to-be-processed feature;

[0023] Matching the feature after filtering processing with the target feature.

[0024] In one of the embodiments, the selecting the feature and the target feature for operation, and selecting a to-be-processed feature corresponding to the target feature based on the operation result, comprises:

[0025] Performing hash operation on the selected feature and the target feature, and determining a monitoring region closest to the target feature based on the operation result;

[0026] Traversing monitoring records belonging to the monitoring region, and selecting a to-be-processed feature with a time gap lower than a preset threshold from the target feature according to the traversal result.

[0027] In one of the embodiments, the establishing a monitoring file corresponding to the current monitoring object in the feature library if the matching fails, comprises:

[0028] If the matching fails, the target feature is stored in the region cache.

[0029] An image feature corresponding to the target feature is acquired, and the feature library is updated asynchronously based on the image feature.

[0030] In one embodiment, the monitoring object processing method further comprises:

[0031] If the matching succeeds, the feature and the feature library in the computing device are updated according to the picture quality of the original monitoring object corresponding to the target feature.

[0032] In one embodiment, if the matching succeeds, the feature and the feature library in the computing device are updated according to the picture quality of the original monitoring object corresponding to the target feature, comprising:

[0033] If the matching succeeds, the target monitoring image is compared with the original monitoring image in the picture quality dimension;

[0034] If the picture quality of the target monitoring image is higher than that of the original monitoring image, the region cache is updated based on the target feature;

[0035] The monitoring object corresponding to the target monitoring image is determined as a new monitoring object, and a monitoring archive corresponding to the new monitoring object is established in the feature library.

[0036] In one embodiment, the monitoring object processing method comprises:

[0037] The features are stored in the region cache in the form of a cache queue.

[0038] In a second aspect, the application further provides a monitoring object processing device based on a cache mechanism. The device comprises:

[0039] A data storage module is configured to receive features corresponding to monitoring objects, and input the features to a region cache for storage, wherein the features are output by a computing device after comparison by a feature library;

[0040] A data receiving module is configured to receive target features corresponding to a current monitoring object output by the computing device;

[0041] A data matching module is configured to match the target features with the stored features.

[0042] An archive establishing module is configured to establish a monitoring archive corresponding to the current monitoring object in the feature library if the matching fails.

[0043] In a third aspect, the present application provides a computer device. The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0044] receiving a feature corresponding to a monitoring object, and inputting the feature into a region cache for storage, the feature being output by a computing device after comparison by a feature library;

[0045] receiving a target feature corresponding to a current monitoring object output by the computing device;

[0046] matching the target feature with the stored feature;

[0047] if the matching fails, establishing a monitoring file corresponding to the current monitoring object in the feature library.

[0048] In a fourth aspect, the present application provides a computer readable storage medium. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the following steps:

[0049] receiving a feature corresponding to a monitoring object, and inputting the feature into a region cache for storage, the feature being output by a computing device after comparison by a feature library;

[0050] receiving a target feature corresponding to a current monitoring object output by the computing device;

[0051] matching the target feature with the stored feature;

[0052] if the matching fails, establishing a monitoring file corresponding to the current monitoring object in the feature library.

[0053] In a fifth aspect, the present application provides a computer program product. The computer program product comprises a computer program, and the computer program is executed by a processor to implement the following steps:

[0054] receiving a feature corresponding to a monitoring object, and inputting the feature into a region cache for storage, the feature being output by a computing device after comparison by a feature library;

[0055] receiving a target feature corresponding to a current monitoring object output by the computing device;

[0056] matching the target feature with the stored feature;

[0057] if the matching fails, establishing a monitoring file corresponding to the current monitoring object in the feature library.

[0058] The aforementioned monitoring object processing method and apparatus based on a caching mechanism adds a local area cache for storing features in the main control program. In subsequent processes, the received target features are compared with the features stored in the area cache to determine whether the current monitored object corresponding to the target feature is a new image feature. By adding a matching process with the area cache, it can prevent missed monitoring objects caused by asynchronous processing within the computing device, which prevents the feature library from being updated in a timely manner. Attached Figure Description

[0059] Figure 1 This is an application environment diagram of a monitoring object processing method based on a caching mechanism in one embodiment;

[0060] Figure 2 This is a flowchart illustrating a monitoring object processing method based on a caching mechanism in one embodiment;

[0061] Figure 3 This is a schematic diagram illustrating the connection relationship of related devices in one embodiment;

[0062] Figure 4 This is a detailed flowchart illustrating a monitoring object processing method based on a caching mechanism in another embodiment;

[0063] Figure 5 This is a structural block diagram of a monitoring object processing device based on a caching mechanism in one embodiment;

[0064] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0066] The monitoring object processing method based on a caching mechanism provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on other network servers. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0067] In one embodiment, such as Figure 2 As shown, a monitoring object processing method based on a caching mechanism is provided, which can be applied to... Figure 1The server in the edge computing server is taken as an example for illustration, including the following steps:

[0068] In step S20, the features corresponding to the monitoring object are received, and the features are input to the region cache for storage. The features are output by the computing device after comparison by the feature library.

[0069] In order to solve the problem of "one person with multiple files" for a single monitoring object due to the asynchronous processing process that cannot update the feature library in time in the prior art, a region cache for storing multiple features is added in the main control program.

[0070] It should be pointed out that the region cache here generally refers to the form of storing features. In actual use, multiple cache queues can be provided in the region cache. Each cache queue can be used to store features sent by a single computing device, or each cache queue can be set to store features sent by multiple computing devices, or multiple cache queues can be set to store features sent by multiple computing devices. The specific storage mode can be adjusted according to actual needs, and is not limited in the embodiment.

[0071] The edge computing server receives a large amount of monitoring images from the snapshot server, and classifies and files them according to the different locations to which the monitoring images belong. After classification and filing, the main control program in the edge computing server transmits the monitoring images to the computing device such as a smart card. After image analysis based on the feature library and feature comparison, the latter outputs features to the main control program, and the main control program stores the received multiple features in the constructed cache queue. The data transmission relationship among the snapshot server, the main control program and the computing device is as shown in Figure 3 .

[0072] A typical computing device includes a smart card. The smart card itself has data processing capability, is connected to a face database, and is connected to the edge computing server through a PCI-E bus, and is specifically used for picture operation and feedback of picture recognition results.

[0073] In step S40, the target features corresponding to the current monitoring object output by the computing device are received.

[0074] After the cache queue is constructed, the main control program of the edge computing server receives the target features corresponding to the current monitoring object output by the computing device after comparison by the feature library in real time, and then matches the target features with the stored features in the cache queue.

[0075] In step S60, the target features are matched with the stored features.

[0076] The received feature of the computing device output is stored in the cache queue, and the newly received target feature is matched with the stored features in the cache queue. The matching manner is to calculate the similarity between the target feature and the stored features in the cache queue. According to the similarity calculation result, it is judged whether the monitoring object corresponding to the target feature is a new monitoring object. If the similarity calculation result is higher than the threshold value, it means that the matching is successful, and it is judged that they are the same monitoring object. If the similarity calculation result is lower than the threshold value, it means that the matching fails, and it is judged that the monitoring object corresponding to the target feature is a new monitoring object that has not been processed.

[0077] In step S80, if the matching fails, a monitoring file corresponding to the current monitoring object is established in the feature library.

[0078] The matching result is divided into two categories. If the matching fails, it means that the monitoring object corresponding to the target feature has not appeared before, and it is a new monitoring object. At this time, the current monitoring object is treated as a new monitoring object in the feature library of the computing device, and a monitoring file corresponding to the current monitoring object is established, so as to facilitate the feature comparison in the subsequent processing process.

[0079] In the above processing process, the main control program locally adds a region cache for storing features, so that the received target feature is compared with the stored features in the region cache to judge whether the current monitoring object corresponding to the target feature is a new image feature. Through the matching processing of the region cache, the monitoring object missing detection caused by the asynchronous processing of the computing device and the failure of the feature library to update in time can be prevented.

[0080] The execution subject of the whole processing method is the main control program in the edge computing server. Although the main control chip running the main control program is not dedicated to floating point operation, since the number of features participating in the comparison here has been reduced by several orders of magnitude, the amount of operation is limited and basically does not affect the original performance of the monitoring image processing.

[0081] In one embodiment, the computing device receives the features of the corresponding monitoring object after comparison based on the feature library, and inputs the features into the region cache in the form of a cache queue, that is, step S20 includes:

[0082] In step S22, a target data structure is obtained, and a region cache is constructed according to the target region structure.

[0083] In step S24, the computing device receives the features of the corresponding monitoring object after comparison based on the feature library.

[0084] In step S26, the features are stored in the region cache in chronological order.

[0085] In implementation, in order to meet the requirement of storing a certain number of features, it is necessary to establish a regional cache at the master program of the edge computing server. Each regional cache is used to receive a queue of monitoring objects of a geographical location. The cache queue is divided by region here, similar to the collection device of the same cell or the same complex being grouped into the same region. It is common sense that a person will not be in multiple regions at the same time, so the queue is divided by region here.

[0086] In order to meet the requirement of storing features, the construction of the regional cache needs to be based on a unified target data structure. After the regional cache is constructed, the features of the corresponding monitoring objects received from the computing device are extracted according to the target data structure, and stored in the regional cache according to the time sequence of the generation of the features.

[0087] In one embodiment, the target data structure is obtained, and the regional cache is constructed according to the target regional structure, that is, step S22 includes:

[0088] Step S222, reading the position information of the monitoring device used to shoot the monitoring object;

[0089] Step S224, obtaining attribute information of the monitoring object;

[0090] Step S226, constructing the target data structure based on the data structure of the position information and the data structure of the attribute information;

[0091] Step S228, constructing the regional cache according to the target data structure.

[0092] In implementation, in the process of using the monitoring device to shoot the monitoring object, the position information of the monitoring device will be carried in the monitoring image, and the attribute information such as the picture format and pixel size of the monitoring image itself can be read from the Exif information of the monitoring image; the target data structure of the regional cache is constructed according to the data structure of the position information and the attribute information.

[0093] In one embodiment, the target feature is matched with the features stored in the form of a cache queue, that is, step S60 includes:

[0094] Step S62, selecting the feature to operate with the target feature, and selecting the to-be-processed feature corresponding to the target feature based on the operation result;

[0095] Step S64, performing time-dimension filtering processing on the to-be-processed feature;

[0096] Step S66, matching the filtered feature with the target feature.

[0097] In implementation, after receiving the target feature, the target feature is matched with the features stored in the cache queue. Considering that a large number of features are stored in each cache queue, in order to shorten the time for matching, the features stored in the cache queue need to be screened.

[0098] The specific screening process is step S62 described above, which includes:

[0099] Step S622, performing hash operation on the selected feature and the target feature, and determining the monitoring area closest to the target feature based on the operation result;

[0100] Step S624, traversing the monitoring records belonging to the monitoring area, and selecting the to-be-processed feature with a time gap lower than a preset threshold from the target feature according to the traversal result.

[0101] In implementation, the number of area caches established in the main control program is large, and typically, each area cache corresponds to one monitoring device. When screening the features, the features in each cache queue need to be matched with the target feature one by one, and then the area cache closest to the target feature is selected according to the hash values of all features in each area cache, and then the monitoring area corresponding to the closest area cache is determined.

[0102] After determining the monitoring area, the features stored in the monitoring area are traversed, and the time difference between each feature obtained by traversal and the target feature is calculated. Only the features with a time difference lower than a preset threshold can reflect that there is a possibility of “one person with multiple records” between the target feature and the object. The selected features need to be processed in subsequent matching.

[0103] In one embodiment, if the matching fails, a monitoring record corresponding to the current monitoring object is established in the feature library, i.e., step S80 includes:

[0104] Step S82, if the matching fails, the target feature is stored in the cache queue;

[0105] Step S84, obtaining the image feature corresponding to the target feature, and updating the feature library asynchronously based on the image feature.

[0106] In implementation, the result of matching failure indicates that the current monitoring object corresponding to the target feature has not appeared, and the information corresponding to the current monitoring object does not exist in the feature library of the computing device. Therefore, two aspects of processing are needed, one is to store the target feature in the cache queue for subsequent matching process, and the other is to update the feature library in the computing device based on the image feature of the current monitoring object corresponding to the target feature.

[0107] It should be noted that the action of updating the feature library is also asynchronous, avoiding blocking the recognition and comparison operations during the feature library updating process. Due to the concurrent processing of the computing device, synchronization of the feature library at this time may cause the same monitored object to be judged as an independent new monitored object multiple times.

[0108] Through the above two aspects of processing, the relevant data of the corresponding new monitored image is increased in the entire two comparison processes of the monitored image, reducing the probability of "one person with multiple files" of the corresponding new monitored image in the subsequent processing process.

[0109] In one embodiment, the monitored object processing method further comprises:

[0110] Step S90, if the matching is successful, updating the feature and the feature library in the computing device according to the picture quality of the original monitored object corresponding to the target feature.

[0111] In implementation, the foregoing step S80 gives the processing step of failing to match the target feature with the feature, and relatively, if the matching is successful, it indicates that the monitored object corresponding to the target feature has appeared before and is not a new monitored object. The feature library in the computing device can be updated according to the monitored image corresponding to the target feature, thereby enhancing the timeliness of the feature library.

[0112] It should be noted that after the matching is successful, a new file is not established, and a refresh operation is performed on the area cache, so that it remains in the area cache for a longer time, facilitating the use in the subsequent matching process.

[0113] Specifically, the operation of updating the feature library is as follows:

[0114] Step S92, if the matching is successful, comparing the target monitored image with the original monitored image in the picture quality dimension;

[0115] Step S94, if the picture quality of the target monitored image is higher than that of the original monitored image, updating the cache queue based on the target feature.

[0116] In implementation, in the state of successful matching, it indicates that the monitored object corresponding to the target monitored image has appeared before, and at this time, the target monitored image needs to be compared with the original monitored image in the picture quality dimension. The purpose of this step is to confirm whether the original monitored image needs to be updated.

[0117] If the picture quality of the target monitored image is higher than that of the original monitored image, in order to improve the accuracy of the subsequent monitored image processing process, the original monitored image needs to be updated. The specific updating operation needs to update both the cache queue and the feature library of the computing device to prevent the occurrence of "one person with multiple files" caused by single updating.

[0118] The embodiment gives the processing scheme of matching the target feature with the stored feature through the foregoing steps. It should be noted that, whether the matching is successful or not, the update of the feature library or the cache queue to the big data monitoring platform is needed after the processing is completed, so as to facilitate subsequent analysis or browsing.

[0119] Based on the foregoing technical scheme, a detailed process of a monitoring object processing method based on a cache mechanism is provided, as shown in Figure 4 , including:

[0120] Step S110, receiving the monitoring image corresponding to the current monitoring object from the snapshot server.

[0121] Step S120, constructing the area cache corresponding to each monitoring device, and transmitting the monitoring image to the area cache corresponding to the monitoring device for storage.

[0122] Step S130, allowing the computing device to analyze the monitoring image based on the feature library to obtain the target feature.

[0123] Step S140, matching the target feature with the stored feature.

[0124] Step S150, if the matching fails, establishing the monitoring file corresponding to the current monitoring object in the feature library.

[0125] Step S160, if the matching succeeds, updating the feature and the feature library in the computing device according to the picture quality of the original monitoring object corresponding to the target feature.

[0126] It should be understood that, although each step in the flowchart involved in each of the above embodiments is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise stated herein, the execution of these steps has no strict sequence limitation, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each of the above embodiments can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0127] Based on the same inventive concept, the embodiment of the present application also provides a cache mechanism-based monitoring object processing device for implementing the cache mechanism-based monitoring object processing method described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more cache mechanism-based monitoring object processing device embodiments provided below can be referred to the limitations of the cache mechanism-based monitoring object processing method described above, which will not be repeated here.

[0128] In one embodiment, as shown in Figure 5 a cache mechanism-based monitoring object processing device 50 is provided, comprising a data storage module 52, a data receiving module 54, a data matching module 56 and an archive establishment module 58, wherein:

[0129] The data storage module 52 is configured to receive the features corresponding to the monitoring object, and input the features to the regional cache for storage, wherein the features are output by the computing device after comparison with the feature library.

[0130] To solve the problem of "one person with multiple archives" for a single monitoring object caused by the asynchronous processing process that cannot update the feature library in time in the prior art, a cache queue for storing multiple features is added in the main control program.

[0131] The edge computing server receives a large amount of monitoring images from the snapshot server, classifies and archives them according to the different locations to which the monitoring images belong. After completing the classification and archiving, the main control program in the edge computing server transmits the monitoring images to the computing device such as a smart card, which outputs the features to the main control program after image analysis based on the feature library and feature comparison. The main control program stores the received multiple features in the constructed cache queue.

[0132] A typical computing device includes a smart card, which has data processing capability, is connected to a face database, and is connected to the edge computing server through a PCI-E bus, and is specifically used for picture operation and feedback of picture recognition results.

[0133] The data receiving module 54 is configured to receive the target features corresponding to the current monitoring object output by the computing device.

[0134] After the cache queue is constructed, the main control program of the edge computing server receives the target features corresponding to the current monitoring object output by the computing device after comparison with the feature library in real time, and then matches the target features with the stored features in the cache queue.

[0135] The data matching module 56 is configured to match the target features with the stored features.

[0136] The received computing device output feature is stored in a cache queue, and the newly received target feature is matched with the stored features in the cache queue. The matching is performed by calculating the similarity between the target feature and the stored features in the cache queue. According to the similarity calculation result, it is determined whether the monitoring object corresponding to the target feature is a new monitoring object. If the similarity calculation result is higher than a threshold, it indicates that the matching is successful, and the two are the same monitoring object. If the similarity calculation result is lower than the threshold, it indicates that the matching fails, and the monitoring object corresponding to the target feature is a new monitoring object that has not been processed.

[0137] The archive establishment module 58 is used to establish a monitoring archive corresponding to the current monitoring object in the feature library if the matching fails.

[0138] The matching result is divided into two categories. If the matching fails, it indicates that the monitoring object corresponding to the target feature has not appeared before and is a new monitoring object. At this time, the current monitoring object is treated as a new monitoring object in the feature library of the computing device, and a monitoring archive corresponding to the current monitoring object is established for subsequent feature comparison.

[0139] In the above processing process, the main control program locally adds a region cache for storing features, so that the received target feature is compared with the stored features in the region cache to determine whether the current monitoring object corresponding to the target feature is a new image feature. Through the matching processing of the region cache, the monitoring object missing detection caused by the asynchronous processing of the computing device and the failure of the feature library to update in time can be prevented.

[0140] The above-mentioned various modules in the monitoring object processing device based on the cache mechanism can be realized by software, hardware, and combinations thereof. The above-mentioned various modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the above-mentioned various modules.

[0141] In one embodiment, a computer device, which can be a server, is provided. The internal structure diagram of the computer device can be as shown in FIG. 1. Figure 6As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as features, target features and feature libraries. The network interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement a monitoring object processing method based on a cache mechanism.

[0142] Those skilled in the art can understand that, Figure 6 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0143] In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the following steps:

[0144] Step S20, receiving the features corresponding to the monitoring object, and inputting the features to the region cache for storage, the features being output by the computing device after comparison by the feature library.

[0145] Among them, in order to solve the problem of "one person with multiple files" for a single monitoring object caused by the asynchronous processing process which cannot update the feature library in time in the prior art, a region cache for storing multiple features is added in the main control program.

[0146] It should be pointed out that the region cache here refers to the form of storing features. In actual use, multiple cache queues can be provided in the region cache, each cache queue can be used to store the features sent by a single computing device, or each cache queue can be set to store the features sent by multiple computing devices, or multiple cache queues can be set to store the features sent by multiple computing devices. The specific storage mode can be adjusted according to actual needs, which is not limited in the present embodiment.

[0147] The edge computing server receives a large number of monitoring images from the snapshot server, and classifies and archives the monitoring images according to different locations to which the monitoring images belong. After the classification and archiving are completed, the main control program in the edge computing server transmits the monitoring images to a computing device such as a smart card. After the computing device performs image analysis based on a feature library and feature comparison, the computing device outputs features to the main control program, and the main control program stores the received features in a constructed cache queue. Figure 3

[0148] The typical computing device includes a smart card, which has data processing capability, is connected to a face database, and is connected to the edge computing server through a PCI-E bus, and is specifically used for picture operation and feedback of picture recognition results.

[0149] In step S40, the target features corresponding to the current monitoring object output by the computing device are received.

[0150] After the cache queue is constructed, the main control program of the edge computing server receives the target features corresponding to the current monitoring object output by the computing device after the feature library comparison, and then matches the target features with the features already stored in the cache queue.

[0151] In step S60, the target features are matched with the stored features.

[0152] The features output by the computing device are stored in the cache queue, and the newly received target features are matched with the features already stored in the cache queue. The matching manner is to calculate the similarity between the target features and the features stored in the cache queue, and to determine whether the monitoring object corresponding to the target features is a new monitoring object according to the similarity calculation result. If the similarity calculation result is higher than a threshold, it indicates that the matching is successful, and the two are determined to be the same monitoring object. If the similarity calculation result is lower than the threshold, it indicates that the matching fails, and the monitoring object corresponding to the target features is a new monitoring object that has not been processed.

[0153] In step S80, if the matching fails, a monitoring file corresponding to the current monitoring object is established in the feature library.

[0154] The matching result is divided into two categories. If the matching fails, it indicates that the monitoring object corresponding to the target features has not appeared before, and is a new monitoring object. At this time, the current monitoring object is treated as a new monitoring object in the feature library of the computing device, and a monitoring file corresponding to the current monitoring object is established, so as to facilitate the feature comparison in the subsequent processing process.

[0155] ​In the above process, the main control program locally adds a region cache for storing features, so that the received target features are compared with the features stored in the region cache to determine whether the current monitoring object corresponding to the target features is a new image feature in the subsequent process. By adding the matching process in the region cache, the occurrence of monitoring object missing detection caused by the failure of the feature library to update in time due to asynchronous processing in the computing device can be prevented.

[0156] In one embodiment, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the following steps are implemented:

[0157] Step S20, receiving the features corresponding to the monitoring object, and inputting the features to the region cache for storage, the features being output by the computing device after comparison by the feature library.

[0158] To solve the problem of "one person with multiple records" for a single monitoring object due to the failure of the feature library to update in time due to asynchronous processing in the prior art, a region cache for storing multiple features is added in the main control program.

[0159] It should be noted that the region cache here refers to the form of storing features. In actual use, multiple cache queues can be provided in the region cache. Each cache queue can be used to store features sent by a single computing device, or each cache queue can be set to store features sent by multiple computing devices, or multiple cache queues can be set to store features sent by multiple computing devices. The specific storage mode can be adjusted according to actual needs, and is not limited in this embodiment.

[0160] The edge computing server receives a large number of monitoring images from the snapshot server, and classifies and archives them according to the different locations to which they belong. After completing the classification and archiving, the main control program in the edge computing server transmits the monitoring images to the computing device of the smart card, for example. After image analysis based on the feature library and feature comparison, the latter outputs the features to the main control program, which stores the received multiple features in the constructed cache queue. The data transmission relationship among the snapshot server, the main control program, and the computing device is as shown in Figure 3 .

[0161] A typical computing device includes a smart card, which has data processing capability, connects to a face database, and is connected to the edge computing server through a PCI-E bus, and is specifically used for picture operation and feedback of picture recognition results.

[0162] Step S40, receiving the target features corresponding to the current monitoring object output by the computing device.

[0163] Wherein, after the construction of the cache queue, the main control program of the edge computing server receives the target features corresponding to the current monitoring object output by the computing device after comparison with the feature library in real time, and then matches the target features with the features stored in the cache queue.

[0164] Step S60, matching the target features with the stored features.

[0165] Wherein, the features output by the computing device are stored in the cache queue, and the newly received target features are matched with the features stored in the cache queue. The specific matching method is to calculate the similarity between the target features and the features stored in the cache queue, and determine whether the monitoring object corresponding to the target features is a new monitoring object according to the similarity calculation result. If the similarity calculation result is higher than the threshold, it means that the matching is successful, and the two are determined to be the same monitoring object. If the similarity calculation result is lower than the threshold, it means that the matching fails, and the monitoring object corresponding to the target features is a new monitoring object that has not been processed.

[0166] Step S80, if the matching fails, a monitoring file corresponding to the current monitoring object is established in the feature library.

[0167] Wherein, the matching result is divided into two categories. If it does not pass the matching, it means that the monitoring object corresponding to the target features has not appeared before, and it is a new monitoring object. At this time, the current monitoring object is treated as a completely new monitoring object in the feature library of the computing device, and a monitoring file corresponding to the current monitoring object is established, so as to facilitate the comparison of features in the subsequent processing process.

[0168] In the above processing process, the main control program locally adds a region cache for storing features, so that the received target features are compared with the features stored in the region cache to determine whether the current monitoring object corresponding to the target features is a new image feature. Through the matching process of the region cache, it can prevent the occurrence of monitoring object missing detection caused by the failure of the feature library to update in time due to asynchronous processing in the computing device.

[0169] In one embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the following steps:

[0170] Step S20, receiving the features corresponding to the monitoring object, and inputting the features to the region cache for storage, the features being output by the computing device after comparison with the feature library.

[0171] Wherein, in order to solve the problem of "one person with multiple files" for a single monitoring object caused by the failure of the feature library to update in time due to asynchronous processing in the prior art, a region cache for storing multiple features is added in the main control program.

[0172] It should be pointed out that the area cache here refers to the form of storing features, and in actual use, multiple cache queues can be provided in the area cache, each of which can be used to store features sent by a single computing device, or each cache queue can be set to store features sent by multiple computing devices, or multiple cache queues can be set to store features sent by multiple computing devices. The specific storage corresponding mode can be adjusted according to actual needs, which is not limited in the embodiment.

[0173] The edge computing server receives a large number of monitoring images from the snapshot server, and classifies and archives them according to the different locations to which the monitoring images belong. After completing the classification and archiving, the main control program in the edge computing server transmits the monitoring images to the computing device such as a smart card. After image analysis based on the feature library and feature comparison, the latter outputs features to the main control program, and the main control program stores the received multiple features in the constructed cache queue. The data transmission relationship among the snapshot server, the main control program and the computing device is as shown in Figure 3 .

[0174] A typical computing device includes a smart card, which has data processing capability, connects to a face database, and is connected to the edge computing server through a PCI-E bus, and is specifically used for picture operation and feedback of picture recognition results.

[0175] Step S40, receiving the target features corresponding to the current monitoring object output by the computing device.

[0176] After the cache queue is constructed, the main control program of the edge computing server receives the target features corresponding to the current monitoring object output by the computing device after comparison with the feature library, and then matches the target features with the features already stored in the cache queue.

[0177] Step S60, matching the target features with the stored features.

[0178] The features output by the computing device are stored in the cache queue, and the newly received target features are matched with the features already stored in the cache queue. The specific matching method is to calculate the similarity between the target features and the features stored in the cache queue, and determine whether the monitoring object corresponding to the target features is a new monitoring object according to the similarity calculation result. If the similarity calculation result is higher than the threshold, it means that the matching is successful, and the two are determined to be the same monitoring object. If the similarity calculation result is lower than the threshold, it means that the matching fails, and the monitoring object corresponding to the target features is a new monitoring object that has not been processed.

[0179] Step S80, if the matching fails, establishing a monitoring file corresponding to the current monitoring object in the feature library.

[0180] The matching result is divided into two categories. If the matching fails, it indicates that the target feature corresponding to the monitored object has not appeared before, and it is a new monitored object. At this time, the current monitored object needs to be added as a new monitored object in the feature library of the computing device, and a monitoring file corresponding to the current monitored object is established, so as to facilitate the feature comparison in the subsequent processing process.

[0181] In the above processing process, the main control program locally adds a region cache for storing features, so that the received target feature is compared with the features stored in the region cache to determine whether the current monitored object corresponding to the target feature is a new image feature in the subsequent process. By adding the matching process in the region cache, the occurrence of the monitoring object missing detection caused by the asynchronous processing of the computing device and the failure of the feature library to update in time can be prevented.

[0182] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are information and data authorized by the user or authorized by all parties.

[0183] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0184] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0185] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A monitoring object processing method based on a caching mechanism, characterized in that, The method for processing the monitored object includes: The system receives the features of the corresponding monitored object, inputs the features into the regional cache for storage, and outputs the features after comparison with the feature library by the computing device. Receive the target features corresponding to the currently monitored object output by the computing device; Match the target feature with the stored features; If the match fails, a monitoring file corresponding to the current monitoring object will be created in the feature library. Matching the target feature with the stored features includes: The stored features are hashed with the target features, and the monitoring area closest to the target features is determined based on the hash result. The monitoring records belonging to the monitoring area are traversed, and features to be processed that have a time difference with the target feature that is lower than a preset threshold are selected based on the traversal results. The features to be processed are then filtered based on the time dimension. The filtered features are matched with the target features.

2. The monitoring object processing method based on a caching mechanism according to claim 1, characterized in that, The step of receiving the features of the corresponding monitored object and inputting the features into the area cache for storage includes: Obtain the target data structure and construct a region cache based on the target data structure; The receiving computing device compares the features of the corresponding monitored object based on a feature database. The features are stored in the region cache in the order they were generated.

3. The monitoring object processing method based on a caching mechanism according to claim 2, characterized in that, The step of obtaining the target data structure and constructing a region cache based on the target data structure includes: Read the location information of the monitoring equipment used to capture images of the monitored objects; Obtain the attribute information of the monitored object; Construct a target data structure based on the data structure of the location information and the data structure of the attribute information; Construct a region cache based on the target data structure.

4. The monitoring object processing method based on a caching mechanism according to claim 1, characterized in that, If a match fails, a monitoring file corresponding to the current monitored object is created in the feature library, including: If a match fails, the target feature is stored in the region cache. Obtain the image features corresponding to the target features, and asynchronously update the feature library based on the image features.

5. The monitoring object processing method based on a caching mechanism according to claim 1, characterized in that, The method for processing monitored objects further includes: If a match is successful, the feature and the feature library in the computing device are updated based on the image quality of the original monitored object corresponding to the target feature.

6. The monitoring object processing method based on a caching mechanism according to claim 5, characterized in that, If a match is successful, the feature and the feature library in the computing device are updated based on the image quality of the original monitored object corresponding to the target feature, including: If a match is found, the target surveillance image is compared with the original surveillance image in terms of image quality. If the image quality of the target surveillance image is higher than that of the original surveillance image, the area cache is updated based on the target features; The monitoring object corresponding to the target monitoring image is identified as a new monitoring object, and a monitoring file corresponding to the new monitoring object is established in the feature library.

7. The monitoring object processing method based on a caching mechanism according to any one of claims 1 to 6, characterized in that, The method for processing the monitored object includes: The features are stored in the region cache as a cache queue.

8. A monitoring object processing device based on a caching mechanism, characterized in that, The monitoring object processing device includes: The data storage module is used to receive the features of the corresponding monitored object, input the features into the area cache for storage, and output the features after comparison with the feature library by the computing device. The data receiving module is used to receive the target features corresponding to the currently monitored object output by the computing device; A data matching module is used to match the target feature with the stored features; The file creation module is used to create a monitoring file corresponding to the current monitoring object in the feature library if the matching fails. Matching the target feature with the stored features includes: The stored features are hashed with the target features, and the monitoring area closest to the target features is determined based on the hash result. The monitoring records belonging to the monitoring area are traversed, and features to be processed that have a time difference with the target feature that is lower than a preset threshold are selected based on the traversal results. The features to be processed are then filtered based on the time dimension. The filtered features are matched with the target features.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the monitoring object processing method based on the caching mechanism according to any one of claims 1 to 7.

Images