Image layout control comparison method, device, computer equipment and storage medium

Abstract

The present application relates to an image control comparison method, device, computer equipment and storage medium. The method includes: obtaining a corrected image set and a control base database feature set; based on the corrected image features in the corrected image set, updating the control base database features in the corresponding control base database feature set, respectively, to obtain at least one archive feature set after the control base database feature set is updated, and the initial control base database features that have not been updated; comparing the acquired captured image features to be detected with the element features in each archive feature set; generating an early warning message when it is detected that the captured image features to be detected are successfully compared with the archive feature set; when it is detected that the captured image features to be detected are failed to be compared with the archive feature set, comparing the captured image features to be detected with the initial control base database features, and generating an early warning message if the comparison is successful. The use of this method can effectively improve the accuracy and efficiency of the control alarm technology.

Description

Technical Field

[0001] The present application relates to the field of control technology, and in particular to an image control comparison method, apparatus, computer equipment, and storage medium. Background Art

[0002] Monitoring refers to the process of monitoring and tracking specific objects. Common monitoring methods include computer monitoring and manual monitoring. Existing manual monitoring methods typically incur high labor and time costs, and the efficiency and accuracy of manual processing are often dependent on the technician's personal status, resulting in relatively unstable accuracy and efficiency. Existing computer monitoring methods often compare facial images in a highly cohesive monitoring library with captured images. If a match is found, the monitoring officer is deemed captured and an alarm is sent. However, this solution places certain demands on factors such as the front-end camera angle and external environment. When the image angle is poor or the quality is low, it is prone to omissions and false alarms.

[0003] Currently, no effective solution has been proposed to address the problems of low accuracy and efficiency of existing control and alarm technologies. Summary of the Invention

[0004] Based on this, it is necessary to provide an image layout control and comparison method, device, computer equipment and storage medium to address the above technical problems.

[0005] In a first aspect, the present application provides an image layout comparison method. The method comprises:

[0006] Obtaining a corrected image set and a control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, thereby obtaining at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated;

[0007] Compare the acquired captured image features to be detected with the element features in each file feature set;

[0008] When it is detected that the features of the captured image to be detected are successfully compared with the archive feature set, an early warning message is generated; when it is detected that the features of the captured image to be detected are failed to be compared with the archive feature set, the features of the captured image to be detected are compared with the features of the initial control base database. If the comparison is successful, an early warning message is generated.

[0009] In one embodiment, obtaining a corrected image set and a control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, respectively, includes:

[0010] Acquire a current initial corrected image set, wherein the initial corrected image set includes at least one initial feature data;

[0011] Perform clustering calculation on the initial feature data in the current initial corrected image set to obtain the current image set, and update the control base database feature set based on the current image set;

[0012] Obtain the next initial corrected image set, and repeat the above steps based on the preset cycle length to periodically update the control base database features in the corresponding control base database feature set.

[0013] In one embodiment, the archive feature set includes at least: a control base database feature for the object to be detected, a main file feature for the object to be detected, sub-file data for the object to be detected, and at least one centroid feature; wherein the main file feature is the feature with the highest similarity to the corresponding control base database feature, the sub-file data is the photographed data for the object to be detected, and the centroid feature is the feature data with the highest similarity to the main file feature for the object to be detected in various time and space environments.

[0014] In one embodiment, the features of the captured image to be detected are compared with the element features in each archive feature set; when it is detected that the features of the captured image to be detected are successfully compared with the archive feature set, an early warning message is generated, including:

[0015] Compare the captured image features to be detected with the control base database features, main file features and centroid features in the archive feature set respectively;

[0016] When it is detected that the number of successful comparisons between the captured image features to be detected and the features in the archive feature set is greater than or equal to a preset warning threshold, a warning message is generated.

[0017] In one embodiment, the acquired captured image features to be detected are compared with the element features in each archive feature set, including:

[0018] Based on the control base database features and all centroid features, weighted processing is performed to obtain the first fitting feature for each file feature set, and the features of the captured image to be detected are compared with the first fitting feature;

[0019] If the comparison fails, weighted processing is performed based on the main file features and all centroid features to obtain a second fitting feature for each file feature set. The captured image features to be detected are compared with the second fitting feature. If the comparison is successful, an early warning message is generated.

[0020] In one embodiment, weighted processing is performed based on the control base database features and all centroid features to obtain a first fitting feature for each archive feature set, including:

[0021] Obtain a preset first weight matrix, and based on the first weight matrix and the corresponding relationship between the control base database feature and the centroid feature, weightedly calculate the first fitting feature of each archive feature set, wherein the control base database feature corresponds to a first weight value in the first weight matrix, the centroid feature corresponds to a second weight value in the first weight matrix, and the first weight value is greater than the second weight value;

[0022] Based on the weighted processing of the main file features and all centroid features, the second fitting features for each file feature set are obtained, including:

[0023] Obtain a preset second weight matrix, and based on the second weight matrix and the correspondence between the main file features and the centroid features, weightedly calculate the second fitting features of each file feature set, wherein the main file features correspond to the third weight value in the second weight matrix, the centroid features correspond to the fourth weight value in the second weight matrix, and the third weight value is greater than the fourth weight value.

[0024] In one embodiment, the acquired captured image features to be detected are compared with the element features in each archive feature set, including:

[0025] Based on the characteristics of the controlled base database, the main file characteristics, and all the centroid characteristics, weighted processing is performed to obtain the third fitting feature for each file feature set. The features of the captured image to be detected are compared with the third fitting feature. If the comparison similarity is detected to be greater than the preset similarity threshold, an early warning information is generated.

[0026] In one embodiment, calculating the fitting feature for each archival feature set includes:

[0027] Obtain a preset third weight matrix, and based on the third weight matrix, and the correspondence between the control base database feature, the main file feature and the centroid feature, weightedly calculate the third fitting feature of each file feature set, wherein the control base database feature corresponds to the fifth weight value in the weight matrix, the main file feature corresponds to the sixth weight value in the weight matrix, and the centroid feature corresponds to the seventh weight value in the weight matrix, the fifth weight value is greater than the sixth weight value, and the sixth weight value is greater than the seventh weight value.

[0028] In a second aspect, the present application also provides an image control and comparison device. The device includes:

[0029] an acquisition module, configured to acquire a corrected image set and a control base database feature set, and update the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, thereby obtaining at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated;

[0030] A comparison module is used to compare the acquired captured image features to be detected with the element features in each file feature set;

[0031] The early warning module is used to generate an early warning message when it is detected that the characteristics of the captured image to be detected are successfully compared with the archival feature set; when it is detected that the characteristics of the captured image to be detected are failed to be compared with the archival feature set, the characteristics of the captured image to be detected are compared with the characteristics in the target feature set, excluding the archival feature set. If the comparison is successful, the early warning message is generated.

[0032] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are performed:

[0033] Obtaining a corrected image set and a control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, thereby obtaining at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated;

[0034] Compare the acquired captured image features to be detected with the element features in each file feature set;

[0035] When it is detected that the features of the captured image to be detected are successfully compared with the archive feature set, an early warning message is generated; when it is detected that the features of the captured image to be detected are failed to be compared with the archive feature set, the features of the captured image to be detected are compared with the features of the initial control base database. If the comparison is successful, an early warning message is generated.

[0036] The above-mentioned image control comparison method, device, computer equipment and storage medium, by correcting the corrected image features in the image set, respectively update the control base database features in the corresponding control base database feature set, and obtain multiple archive feature sets after the control base database feature set is updated, and the initial control base database features that have not been updated; the acquired captured image features to be detected are compared with the element features in each archive feature set; when it is detected that the captured image features to be detected are successfully compared with the archive feature set, an early warning message is generated; when it is detected that the captured image features to be detected are failed to be compared with the archive feature set, the captured image features to be detected are compared with the initial control base database features, and if the comparison is successful, an early warning message is generated. The data quality of the control base database features can be improved through this application, and the captured image features to be detected are first compared with the archive features and then compared with the initial control base database features, which can effectively improve the accuracy of the comparison results and reduce problems such as missed comparisons and wrong comparisons. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A diagram showing an application environment of an image control and comparison method in one embodiment;

[0038] Figure 2 1 is a flow chart of an image control and comparison method according to an embodiment;

[0039] Figure 3 A schematic diagram of the layout of an archive feature set structure in one embodiment;

[0040] Figure 4 Schematic diagram of the flow of an image control and comparison method in a preferred embodiment;

[0041] Figure 5 2. It is a structural block diagram of an image control and comparison device in one embodiment;

[0042] Figure 6 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0044] The image layout control comparison method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. Obtain a corrected image set and a control base database feature set, and based on the corrected image features in the corrected image set, update the control base database features in the corresponding control base database feature set respectively to obtain multiple archive feature sets after the control base database feature set is updated, and the initial control base database features that have not been updated; compare the acquired captured image features to be detected with the element features in each archive feature set, and generate a warning message when it is detected that the captured image features to be detected are successfully compared with the archive feature set. When it is detected that the captured image features to be detected are failed to be compared with the archive feature set, compare the captured image features to be detected with the initial control base database features. If the comparison is successful, generate a warning message. Terminal 102 may include, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices may include smart speakers, smart TVs, smart air conditioners, and smart car devices. Portable wearable devices may include smart watches, smart bracelets, and head-mounted devices. Server 104 may be implemented as a standalone server or a server cluster consisting of multiple servers.

[0045] In one embodiment, Figure 2 As shown, an image layout comparison method is provided, which is applied to Figure 1 The following steps are used as an example to illustrate the server in the example:

[0046] Step S202: obtain a corrected image set and a control base database feature set, and based on the corrected image features in the corrected image set, update the control base database features in the corresponding control base database feature set, and obtain at least one archive feature set after the control base database feature set is updated, and the initial control base database feature that has not been updated.

[0047] This application is often used in scenarios involving the detection or tracking of one or more controlled objects, including but not limited to personnel and vehicles. The aforementioned control base database is generally a pre-set image base database for the controlled objects, such as frontal photos or ID photos of the controlled personnel, or the license plates and appearance features of the controlled vehicles. In summary, feature extraction is performed on the pre-set image base database of the controlled objects to obtain the aforementioned control base database feature set. The above-mentioned corrected image set is a set of images that can be used to update the control base database feature set. The method for obtaining the corrected image set includes taking a snapshot of the control object. It can be understood that the method for obtaining the corrected image set is not limited in this embodiment. All methods for obtaining corrected images that can be used to update the control base database feature set should fall within the scope of protection of this application. Preferably, in actual applications, the above-mentioned corrected image set can be image data of the control object captured by an image acquisition device such as a camera. The control base database is updated based on the corrected image set actually captured, effectively avoiding the influence of factors such as the more complex angles and environments in which the control object is actually photographed, which reduces the accuracy of subsequent comparison results. Furthermore, the control base database feature set should be pre-acquired or pre-established based on the control object; and the corrected image set is obtained after the control base database feature set is established, and is used to update and correct the control base database feature set.

[0048] The corrected image set includes multiple corrected image features. These features are categorized by controlled object, meaning each controlled object corresponds to at least one corrected image feature. Multiple different corrected image features corresponding to the same controlled object include, but are not limited to, images of the controlled object captured from different angles and in different scenarios. Similarly, the control base database feature set includes multiple control base database features, each of which is a preset image feature corresponding to a controlled object.

[0049] After obtaining the corrected image set and the control base database feature set, the corresponding control base database features are updated based on the corrected image features to obtain an updated archive feature set. The updating method for the control base database features includes, but is not limited to, merging the corrected image features with the control base database features, i.e., the updated archive feature set includes {initial control base database features, corrected image features}; or replacing the corresponding non-updated initial control base database features with higher-quality images from the corrected image features, etc. In actual applications, the obtained corrected image set is generally unable to update all control base database features. For example, the preset control base database feature set includes base database features of 100 control objects, but the obtained corrected image set only includes corrected image features of 10 of these control objects. Therefore, after updating the control base database features using the corrected image features, 10 archive feature sets corresponding to the corrected image features and 90 non-updated initial control base database features of the control objects are usually obtained.

[0050] Step S204 : comparing the acquired captured image features to be detected with the element features in each file feature set.

[0051] Among them, each archival feature set includes multiple element features. Different element features in the same archival feature set reflect different features of the same controlled object, where the element features include but are not limited to image features of the controlled object captured in different scenes, image features of the controlled object at different angles, etc. After obtaining the captured image features to be detected, the captured image features to be detected are compared with the element features. It is understandable that the acquisition of the above-mentioned captured image features to be detected includes but is not limited to extracting the edges, corners, and textures of the object to be detected in the image, or extracting image features based on deep learning methods, etc.

[0052] Step S206, when it is detected that the features of the captured image to be detected are successfully compared with the archive feature set, an early warning message is generated; when it is detected that the features of the captured image to be detected are failed to be compared with the archive feature set, the features of the captured image to be detected are compared with the features of the initial control base database. If the comparison is successful, an early warning message is generated.

[0053] Among them, the features of the captured image to be detected are first compared with the archive feature set. If the comparison is successful, an early warning is issued, indicating that the control object has been detected. If the comparison fails, it is compared with the initial control base database features. Similarly, if the comparison is successful, an early warning is issued. Through steps S202 to S206, the control base database features are updated based on the corrected image set, and are compared with the archive feature set first during the comparison. This can achieve automatic expansion of the control image features of a single angle scene to multiple capture angles, multiple quality levels, and multiple spatiotemporal domains, reducing the quality requirements of the captured image to be detected, reducing the probability of missed comparisons and false comparisons, and improving the accuracy of control comparisons.

[0054] In some embodiments, the method further comprises:

[0055] Acquire a current initial corrected image set, wherein the initial corrected image set includes at least one initial feature data;

[0056] Perform clustering calculation on the initial feature data in the current initial corrected image set to obtain the current image set, and update the control base database feature set based on the current image set;

[0057] Obtain the next initial corrected image set, and repeat the above steps based on the preset cycle length to periodically update the control base database features in the corresponding control base database feature set.

[0058] Specifically, an initial corrected image set is obtained, wherein the initial corrected image set is generally obtained by continuously capturing real-time images and their spatiotemporal domain (time, spatial location), external environment, etc., wherein the initial feature data reflects the characteristics of different captured images. After obtaining the current initial corrected image set, the current initial corrected image set is clustered and archived to obtain the current image set, wherein the current initial corrected image set is clustered into clusters, and similar clusters are merged into a large archive set to obtain the above-mentioned current image set, and the clustering algorithm is used again based on the current image set and the control base database feature set to form an updated cluster set. Repeat the above method, and periodically use the merging algorithm to merge similar categories of the cluster set to form the above-mentioned archive feature set. In some preferred embodiments, each file feature set includes {main file feature, k centroid features, sub-file data set}, wherein the main file feature represents the best target angle of the captured image, wherein the determination of the main file feature can be determined according to the actual situation, such as the main file feature can be obtained based on the image with the most captured information of the control object; or the main file feature can be obtained based on the captured image closest to the control base image, etc.; the above k centroid features represent the feature data with the highest similarity to the main file feature in various spatiotemporal environments, wherein the spatiotemporal environment includes the spatiotemporal domain information, angle information of the object to be detected, and the like. , scene information, etc., that is, the above-mentioned centroid feature represents, in at least one spatiotemporal domain, the same target associated with the main file has higher quality or the highest quality images at different angles and in different scenes, wherein different angles include the front, side, back and other angles of the controlled object, and different scenes include rainy days, foggy days, daytime, night and other scenes; the above-mentioned sub-file data set records the snapshot information of the controlled object; the quality judgment can be determined by relevant technical personnel according to actual needs, and preferably, feature data with a high similarity to the main file feature (such as a similarity greater than 60%, or 80%, etc.) can be determined as a higher quality centroid feature. Through this embodiment, the control base database feature set can be periodically and automatically updated to enrich the archive feature set to improve the accuracy of control comparison.

[0059] In some embodiments, the archive feature set includes at least: a control base database feature for the object to be detected, a main file feature for the object to be detected, sub-file data for the object to be detected, and at least one centroid feature; wherein the main file feature is the feature with the highest similarity to the corresponding control base database feature, the sub-file data is the photographed data for the object to be detected, and the centroid feature is the feature data with the highest similarity to the main file feature for the object to be detected in various time and space environments.

[0060] Specifically, Figure 3The figure is a schematic diagram of the structural layout of the archive feature set in one embodiment. Each archive unit in the set includes a control base database feature, a main file feature, multiple centroid features, and sub-file data. In this embodiment, the feature with the highest similarity to the control base database feature is used as the main file feature. Furthermore, in some preferred embodiments, the above-mentioned archive feature set also includes a spatiotemporal attribute set. The spatiotemporal attribute set reflects the spatiotemporal information of the corresponding image when it is acquired. It is understandable that when comparing with the captured image feature to be detected, the spatiotemporal information of the two also needs to be compared.

[0061] In some embodiments, the method further comprises:

[0062] Compare the captured image features to be detected with the control base database features, main file features and centroid features in the archive feature set respectively;

[0063] When it is detected that the number of successful comparisons between the captured image features to be detected and the features in the archive feature set is greater than or equal to a preset warning threshold, a warning message is generated.

[0064] Specifically, each archive feature set includes at least one element feature, which includes but is not limited to the control base warehouse feature, main file feature and centroid feature mentioned above. The features of the captured image to be detected are compared one by one with the above element features. If the number of successful comparisons is greater than the preset warning threshold, it means that the comparison between the captured image to be detected and the archive feature set is successful, and a warning message is generated at this time, wherein the above warning threshold is preferably 1, that is, an alarm is issued when one of the control base warehouse feature, main file feature, and multiple centroid features is compared with the capture image feature to be detected. This embodiment can improve the problems of missed comparison, wrong comparison, etc. caused by comparing the captured image with the base warehouse feature of a single angle when the captured image has a poor angle, insufficient clarity, or low quality, etc., reduce the quality requirements for the captured image, and improve the accuracy of the control comparison.

[0065] In some embodiments, the method further comprises:

[0066] Based on the control base database features and all centroid features, weighted processing is performed to obtain the first fitting feature for each file feature set, and the features of the captured image to be detected are compared with the first fitting feature;

[0067] If the comparison fails, weighted processing is performed based on the main file features and all centroid features to obtain a second fitting feature for each file feature set. The captured image features to be detected are compared with the second fitting feature. If the comparison is successful, an early warning message is generated.

[0068] Specifically, in this embodiment, the first fitting feature is first calculated. The first fitting feature is obtained by weighted processing based on the control base database feature and all centroid features, and the first fitting feature is compared with the feature of the captured image to be detected. If the similarity between the two is higher than a preset threshold, an alarm is issued, wherein the threshold can be set to 60% or 80%, etc.; and if the similarity is not higher than the threshold, it is further compared with the second fitting feature. The second fitting feature is obtained by weighted processing based on the main file feature and the centroid feature. If the similarity between the feature of the captured image to be detected and the second fitting feature is higher than the threshold, an early warning message is generated, otherwise no early warning is issued. In this embodiment, the centroid feature can effectively improve the accuracy of the comparison of some captured images to be detected with incorrect angles and poor quality, and the weighted calculation with the control base database feature or the main file feature can effectively reduce false alarms.

[0069] In some embodiments, weighted processing is performed based on the control base database features and all centroid features to obtain a first fitting feature for each archive feature set, including:

[0070] Obtain a preset first weight matrix, and based on the first weight matrix and the corresponding relationship between the control base database feature and the centroid feature, weightedly calculate the first fitting feature of each archive feature set, wherein the control base database feature corresponds to a first weight value in the first weight matrix, the centroid feature corresponds to a second weight value in the first weight matrix, and the first weight value is greater than the second weight value;

[0071] Based on the weighted processing of the main file features and all centroid features, the second fitting features for each file feature set are obtained, including:

[0072] Obtain a preset second weight matrix, and based on the second weight matrix and the correspondence between the main file features and the centroid features, weightedly calculate the second fitting features of each file feature set, wherein the main file features correspond to the third weight value in the second weight matrix, the centroid features correspond to the fourth weight value in the second weight matrix, and the third weight value is greater than the fourth weight value.

[0073] Specifically, in this embodiment, the first fitting feature is first calculated according to the first weight matrix, wherein the first weight matrix can be defined as a (k+1) row × 1 column weight matrix W (k+1)x1 ={W s W0 W1 … W k-1} T , where W s is the weight corresponding to the control base database feature, that is, the first weight value mentioned above, W0 to W k-1 is the weight corresponding to the centroid feature, that is, the second weight value mentioned above, which is the weight of each k centroid, and the first weight value is greater than the second weight value. In summary, the calculation of the first fitting feature f is:

[0074] f=S i ×W s +centroid[0]×W0+ … centroid[k-1]×W k-1

[0075] Where Si is the above-mentioned control base database feature, centroid[0]…centroid[k-1] is the centroid feature of K snapshots in different spatiotemporal environments. According to the above method, the first fitting feature of each archive feature set can be calculated.

[0076] Similarly, the features of the captured image to be detected are first compared with the first fitting features. If the comparison fails, the features are compared with the second fitting features. In this embodiment, the second fitting features are calculated based on the second weight matrix. The second weight matrix can be defined as a (k+1) row × 1 column weight matrix W (k+1)x1 ={W m W0 W1 … W k-1} T , where W m The main feature weight, that is, the third weight value, W0 to W k-1 is the weight corresponding to the centroid feature, i.e., the fourth weight value mentioned above, and wherein the third weight value is greater than the fourth weight value. Further, the first weight value mentioned above may be equal to or different from the third weight value, which can be determined by relevant technical personnel. Similarly, the second weight value may be equal to or different from the fourth weight value. Preferably, the second weight value is equal to the fourth weight value.

[0077] In summary, the calculation of the second fitting feature f' is:

[0078] f'=P i ×W m +centroid[0]×W0+ … centroid[k-1]×W k-1

[0079] Among them, P i is the main file feature. Similarly, centroid[0]…centroid[k-1] are the centroid features of K different spatiotemporal environments. According to the above method, the second fitting feature of each file feature set can be calculated.

[0080] Through this embodiment, by setting weights, the main file features and the control base database features are mainly used, and the centroid features are used as a supplement to calculate the fitting features for comparison with the object to be detected and captured. While ensuring the accuracy of the comparison, the probability of missed reports and false alarms is reduced.

[0081] In some embodiments, the method further comprises:

[0082] Based on the characteristics of the controlled base database, the main file characteristics, and all the centroid characteristics, weighted processing is performed to obtain the third fitting feature for each file feature set. The features of the captured image to be detected are compared with the third fitting feature. If the comparison similarity is detected to be greater than the preset similarity threshold, an early warning information is generated.

[0083] Specifically, in this embodiment, a weighted calculation is performed on the monitored base inventory features, the main file features, and the centroid features to obtain the aforementioned third fitting feature. This third fitting feature is then compared with the features of the captured image to be detected. If the comparison similarity exceeds a similarity threshold, an alert is generated. The similarity threshold can be 60%. In this embodiment, the weighted comparison of these three features creates a strict standard, effectively reducing the possibility of false alarms.

[0084] In some embodiments, the method further comprises:

[0085] Obtain a preset third weight matrix, and based on the third weight matrix, and the correspondence between the control base database feature, the main file feature and the centroid feature, weightedly calculate the third fitting feature of each file feature set, wherein the control base database feature corresponds to the fifth weight value in the weight matrix, the main file feature corresponds to the sixth weight value in the weight matrix, and the centroid feature corresponds to the seventh weight value in the weight matrix, the fifth weight value is greater than the sixth weight value, and the sixth weight value is greater than the seventh weight value.

[0086] Specifically, a third weight matrix can be preset first. The elements in the weight matrix are all weight values ​​corresponding to different features, including weight matrices corresponding to the control base warehouse feature, main file feature and centroid feature respectively. The third weight matrix can be set as a (k+2) row × 1 column weight matrix W (k+2)x1 = {w s , w m , w0, w1… w k-1} T , where w s is the weight of the control base feature, i.e. the fifth weight value mentioned above, which has the highest weight; w m is the main feature weight, i.e. the sixth weight value mentioned above, which is second only to the fifth weight value; w1…w k-1 is the seventh weight value, which is the lowest, and is the weight of k centroids. The calculation formula of the third fitting feature f'' is as follows:

[0087] f''= s i × w s + p i × w m+ centroid[0] × w0 + … + centroid[k-1] ×w k-1

[0088] The features of the captured image to be detected are compared with the third fitting features. If the similarity is greater than a preset similarity threshold and the time and space environments are the same, it is considered a match and a corresponding warning information is generated.

[0089] This application also provides a preferred embodiment of an image layout control comparison method, Figure 4 Schematic diagram of the flow of the image control and comparison method in a preferred embodiment.

[0090] Step S410: Obtain the control base database feature set. Take the image material of the control object as the base database, extract the control object features based on the control image material and the material target structured attribute information, and form the control image feature set S = {s i | 1≤i≤m}, thus obtaining the above-mentioned control base database feature set.

[0091] Step S420: Obtain an initial set of corrected images. In this embodiment, the corrected image set is a collection of snapshot images of various scenes captured by various front-end camera devices. The captured images are extracted using relevant algorithms to obtain features and attributes. The initial set of corrected images includes the captured image features and attribute information corresponding to the image features, such as target angle, time, weather, and spatial location.

[0092] Step S430: Cluster and archive the initial corrected image set. Cluster the captured initial corrected image set into clusters, merge similar clusters into large archive sets, and form a set of multiple target archives. For the captured image features and the historical captured image feature sets, use a clustering algorithm to form a cluster set P t , periodically use the combined algorithm to compare with the cluster set P t Similar categories are combined to form a modified image set P, P={p j | 0<j≤n}, each file unit p j The dataset consists of {main file features, k centroid features, and sub-file datasets}. The main file represents the best captured image angle, and the k centroid features represent the centroid features of the best k images from different angles, including front, side, and back scenes, as well as different scenes such as daytime, nighttime, rainy, and foggy. The merged archive set P is pushed to the aforementioned control base database feature set, and subsequent captured images are clustered and archived, repeating the previous process.

[0093] Step S440: Update the control base database feature set based on the modified image set. i| 1≤i≤m} as the control image feature set and periodically receive the clustered corrected image set P={p j | 0<j≤n}, for each file p in P j With s in the bottom library i Compare them one by one, if they match, the file is considered p j To control the base members i If the file is not matched, skip it and select the next file p j+1 Continue to compare until all the profile feature sets are traversed. After the update is completed, the updated SP set is obtained, where each SP i Elements include: {original control base database feature S i 、Main file feature p i , k centroid features, sub-file data, spatiotemporal domain attribute sets} and other field information.

[0094] Step S450 compares the captured image features to be detected with the first fitting features. The first fitting features are weighted calculations based on a preset first weight matrix, the corresponding relationships between the control base database features and the centroid features, for each archive feature set. If a match is found, a warning message is generated; if not, the process proceeds to step S460.

[0095] Step S460 compares the captured image features to be detected with the second fitting features. The second fitting features are weighted calculations based on a preset second weight matrix and the correspondence between the primary file features and the centroid features for each file feature set. If a match is found, a warning message is generated. If not, the process proceeds to step S470.

[0096] Step S470: compare the captured image features to be detected with the unupdated initial control base database features. If the comparison is successful, an early warning message is generated. The initial control base database features are those that have not been updated by the modified image set P. i .

[0097] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0098] Based on the same inventive concept, the embodiments of the present application also provide an image control and comparison device for implementing the aforementioned image control and comparison method. The implementation solution provided by this device is similar to the implementation solution described in the aforementioned method. Therefore, the specific limitations of one or more of the following embodiments of the image control and comparison device can be found in the above-mentioned limitations of the image control and comparison method, and will not be repeated here.

[0099] In one embodiment, Figure 5 As shown, an image control and comparison device is provided, comprising: an acquisition module 51, a comparison module 52 and an early warning module 53, wherein:

[0100] An acquisition module 51 is configured to acquire a corrected image set and a control base database feature set, and based on the corrected image features in the corrected image set, update the control base database features in the corresponding control base database feature set, thereby obtaining at least one updated archive feature set of the control base database feature set and an initial control base database feature that has not been updated.

[0101] A comparison module 52 is used to compare the acquired captured image features to be detected with the element features in each file feature set;

[0102] The early warning module 53 is used to generate an early warning message when it is detected that the features of the captured image to be detected are successfully compared with the archive feature set; when it is detected that the features of the captured image to be detected are failed to be compared with the archive feature set, the features of the captured image to be detected are compared with the features in the target feature set except the archive feature set. If the comparison is successful, an early warning message is generated.

[0103] Each module in the aforementioned image layout control and comparison device can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a computer device memory in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0104] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory and a network interface connected via a system bus. 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 image control and comparison related data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, an image control and comparison method is implemented.

[0105] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0106] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0107] Obtaining a corrected image set and a control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, thereby obtaining at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated;

[0108] Compare the acquired captured image features to be detected with the element features in each file feature set;

[0109] When it is detected that the features of the captured image to be detected are successfully compared with the archive feature set, an early warning message is generated; when it is detected that the features of the captured image to be detected are failed to be compared with the archive feature set, the features of the captured image to be detected are compared with the features of the initial control base database. If the comparison is successful, an early warning message is generated.

[0110] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0111] Obtaining a corrected image set and a control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, thereby obtaining at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated;

[0112] Compare the acquired captured image features to be detected with the element features in each file feature set;

[0113] When it is detected that the features of the captured image to be detected are successfully compared with the archive feature set, an early warning message is generated; when it is detected that the features of the captured image to be detected are failed to be compared with the archive feature set, the features of the captured image to be detected are compared with the features of the initial control base database. If the comparison is successful, an early warning message is generated.

[0114] 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 used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0115] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), data processing logic devices based on quantum computing, and the like.

[0116] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. An image control comparison method, characterized in that: The method comprises: Acquire a corrected image set and a control base database feature set, and based on the corrected image features in the corrected image set, update the control base database features in the corresponding control base database feature set, respectively, to obtain at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated; wherein, the archive feature set at least includes: a control base database feature for an object to be detected, a main file feature for an object to be detected, sub-file data for an object to be detected, and at least one centroid feature; wherein, the main file feature is a feature with the highest similarity to the corresponding control base database feature, the sub-file data is photographed data for the object to be detected, and the centroid feature is feature data with the highest similarity to the main file feature for the object to be detected in various spatiotemporal environments; wherein, the control base database is a preset image base database for the control object, and feature extraction is performed on the image base database to obtain the control base database feature set; the corrected image set is a set of images for updating the control base database feature set, and the corrected image set includes multiple corrected image features, and the corrected image features are based on the control object; Wherein, the obtaining of the corrected image set and the control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, respectively, includes: obtaining a current initial corrected image set, wherein the initial corrected image set includes at least one initial feature data; performing clustering calculation on the initial feature data in the current initial corrected image set to obtain a current image set, and updating the control base database feature set based on the current image set; obtaining a next initial corrected image set, and repeating the above steps based on a preset period duration to periodically update the control base database features in the corresponding control base database feature set; Comparing the acquired captured image features to be detected with the element features in each of the archival feature sets; wherein each of the archival feature sets includes multiple element features, and different element features in the same archival feature set reflect different features of the same monitored object; When it is detected that the features of the captured image to be detected are successfully compared with the set of archival features, an early warning message is generated; when it is detected that the features of the captured image to be detected are failed to be compared with the set of archival features, the features of the captured image to be detected are compared with the features of the initial control base database. If the comparison is successful, the early warning message is generated.

2. The method according to claim 1, characterized in that The method of comparing the acquired captured image features to be detected with the element features in each of the archive feature sets; generating an early warning message when it is detected that the captured image features to be detected are successfully compared with the archive feature set, includes: Comparing the captured image features to be detected with the control base database features, the main file features and the centroid features in the archive feature set respectively; When it is detected that the number of successful comparisons between the captured image features to be detected and the features in the archive feature set is greater than or equal to a preset warning threshold, the warning information is generated.

3. The method according to claim 1, characterized in that The step of comparing the acquired captured image features to be detected with the element features in each of the archive feature sets includes: Performing weighted processing based on the control base database features and all the centroid features to obtain a first fitting feature for each of the archive feature sets, and comparing the captured image features to be detected with the first fitting feature; If the comparison fails, weighted processing is performed based on the main file features and all the centroid features to obtain a second fitting feature for each of the file feature sets, and the captured image features to be detected are compared with the second fitting features. If the comparison is successful, the warning information is generated.

4. The method according to claim 3, characterized in that The weighted processing based on the control base database feature and all the centroid features to obtain the first fitting feature for each of the archive feature sets includes: Obtaining a preset first weight matrix, and weightedly calculating the first fitting feature for each of the archive feature sets based on the first weight matrix and the corresponding relationship between the control base database feature and the centroid feature, wherein the control base database feature corresponds to a first weight value in the first weight matrix, the centroid feature corresponds to a second weight value in the first weight matrix, and the first weight value is greater than the second weight value; The weighted processing based on the main file feature and all the centroid features to obtain the second fitting feature for each of the file feature sets includes: Obtain a preset second weight matrix, and based on the second weight matrix and the correspondence between the main file features and the centroid features, weightedly calculate the second fitting features of each of the file feature sets, wherein the main file features correspond to the third weight value in the second weight matrix, the centroid features correspond to the fourth weight value in the second weight matrix, and the third weight value is greater than the fourth weight value.

5. The method according to claim 1, wherein The step of comparing the acquired captured image features to be detected with the element features in each of the archive feature sets includes: Based on the control base database features, the main file features, and all the centroid features, weighted processing is performed to obtain a third fitting feature for each of the file feature sets. The captured image features to be detected are compared with the third fitting features. If the comparison similarity is detected to be greater than a preset similarity threshold, the warning information is generated.

6. The method according to claim 5, characterized in that The weighted processing based on the control base database feature, the main file feature, and all the centroid features to obtain the third fitting feature for each of the file feature sets includes: Obtain a preset third weight matrix, and based on the third weight matrix and the correspondence between the control base database feature, the main file feature and the centroid feature, weightedly calculate the third fitting feature of each of the archive feature sets, wherein the control base database feature corresponds to the fifth weight value in the weight matrix, the main file feature corresponds to the sixth weight value in the weight matrix, and the centroid feature corresponds to the seventh weight value in the weight matrix, the fifth weight value is greater than the sixth weight value, and the sixth weight value is greater than the seventh weight value.

7. An image control and comparison device, characterized in that: The device comprises: An acquisition module is used to acquire a corrected image set and a control base database feature set, and based on the corrected image features in the corrected image set, respectively update the control base database features in the corresponding control base database feature set to obtain at least one archive feature set after the control base database feature set is updated, and an initial control base database feature that has not been updated; wherein, the archive feature set at least includes: control base database features for the object to be detected, main file features for the object to be detected, sub-file data for the object to be detected, and at least one centroid feature; wherein, the main file feature is the feature that corresponds to the control base database. The feature with the highest similarity to the control base database feature, the sub-file data is the shooting data for the object to be detected, and the centroid feature is the feature data with the highest similarity to the main file feature for the object to be detected in various time and space environments; wherein the control base database is a preset image base database for the control object, and feature extraction is performed on the image base database to obtain the control base database feature set; the corrected image set is a set of images that update the control base database feature set, and the corrected image set includes multiple corrected image features, and the corrected image features are based on the control object; Wherein, the obtaining of the corrected image set and the control base database feature set, and updating the control base database features in the corresponding control base database feature set based on the corrected image features in the corrected image set, respectively, includes: obtaining a current initial corrected image set, wherein the initial corrected image set includes at least one initial feature data; performing clustering calculation on the initial feature data in the current initial corrected image set to obtain a current image set, and updating the control base database feature set based on the current image set; obtaining a next initial corrected image set, and repeating the above steps based on a preset period duration to periodically update the control base database features in the corresponding control base database feature set; a comparison module for comparing the acquired captured image features to be detected with the element features in each of the archival feature sets; wherein each of the archival feature sets includes multiple element features, and different element features in the same archival feature set reflect different features of the same monitored object; The early warning module is used to generate an early warning message when it is detected that the characteristics of the captured image to be detected are successfully compared with the archival feature set; when it is detected that the characteristics of the captured image to be detected are failed to be compared with the archival feature set, the characteristics of the captured image to be detected are compared with the characteristics in the target feature set, excluding the archival feature set. If the comparison is successful, the early warning message is generated.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

Images