Method and system for judging availability of intelligent camera

Abstract

The invention discloses an intelligent camera availability judgment method and system, and relates to the technical field of intelligent prevention and control, and the method comprises the steps: calculating the online rate of a single intelligent camera in an analysis time period; counting the snapshot amount of daily snapshot data of a single intelligent camera; calculating the total effective snapshot rate of the single intelligent camera; forming a target recognition rate of a single intelligent camera in an analysis time period; forming a combat generation rate of a single intelligent camera in an analysis time period; establishing an intelligent camera availability rate index, and screening the intelligent cameras of which the availability rates do not meet requirements through dynamic division of the availability rates of the cameras to carry out graded early warning; the actual operation condition is fully considered, an intelligent camera early warning integral system is constructed, integral conversion is carried out on early warning, and finally clustering visual display of integral cameras is achieved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent prevention and control technology, and in particular to a method and system for determining the availability of an intelligent camera. Background Art

[0002] With the development of video surveillance technology, cameras have covered every corner of the city, including urban public areas, enterprises, institutions, communities, hospitals, schools and other places, becoming the guardian of public safety. Whether it is a video surveillance system with tens of thousands or dozens of cameras, continuous management and maintenance of cameras are required to play a lasting role. The traditional approach is mainly to monitor the offline status of cameras through a system, conduct real-time warnings and offline time statistics for offline cameras, so as to detect camera failures and deal with them, and facilitate statistical assessment of camera online rates.

[0003] At this stage, with the development of video AI technology, the proportion of intelligent cameras is gradually increasing. Moreover, with the upgrading of video surveillance systems, intelligent cameras will achieve comprehensive coverage. Compared with ordinary cameras, they have the ability to recognize, capture and analyze targets in video images, realize the pre-analysis capability, and thus fully tap the valuable information in the video.

[0004] However, it is not enough to only use online monitoring in the existing technology for such cameras. It is impossible to fully grasp their availability and it is difficult to make comprehensive judgments on their multiple dimensions. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for determining the availability of an intelligent camera to solve the following technical problems: How to comprehensively judge the availability of intelligent cameras from multiple dimensions.

[0006] The purpose of the present invention can be achieved through the following technical solutions: In a first aspect, the present invention discloses a method for determining the availability of an intelligent camera, which sequentially performs the following steps: S1: Obtain the online and offline status of a specified smart camera during the analysis period, and calculate the online rate of a single smart camera during the analysis period; S2: count the number of captured data of a single smart camera every day; S3: Evaluate the capture quality of the captured data, count the effective capture volume of a single smart camera per day and the overall effective capture volume in the analysis period, and calculate the overall effective capture rate of a single smart camera; S4: Analyze the target recognition status of the effective captured data to form the target recognition rate of a single smart camera within the analysis time period; S5: Count the business performance data of the specified smart camera to form the performance rate of a single smart camera within the analysis period; S6: Establish an availability index for smart cameras, and screen out smart cameras whose availability does not meet the requirements for early warning; S7: Set different scores for different levels of warnings, build a smart camera warning points system, and realize cluster display of scoring cameras.

[0007] In a further solution of the present invention: in step S1, the online rate of a single smart camera in the analysis time period is calculated as follows: S11: count the offline time of a single smart camera; S12: converting the total duration of the analysis period into the expected online duration, and obtaining the actual online duration of the smart camera in the analysis period by the difference between the expected online duration and the offline duration; S13: Calculate the ratio of the actual online duration to the expected online duration to obtain the online rate of a single smart camera.

[0008] In a further solution of the present invention: in step S2, the captured data of the single smart camera includes the camera number, camera name, target ID, capture time, picture URL, attribute information, and quality score.

[0009] In a further solution of the present invention: in step S2, the daily capture amount of the single smart camera is calculated as follows: S21: Taking a single smart camera as a unit, obtaining its captured data by camera number; S22: Filter and retain all captured data of the smart camera within one day according to the capture time; S23: Filter the repeated captured data according to the target ID, and perform data filtering; identify the captured data with unavailable image URLs, and perform data filtering; identify the captured data with large capture time deviations, and perform data filtering; S24: After excluding the filtered data in step S23, the daily capture volume of a single smart camera is counted.

[0010] In a further solution of the present invention: in step S3, the overall effective capture rate of the single smart camera is calculated as follows: S31: Setting the picture quality benchmark score; S32: Compare the quality scores of the daily captured data in step S2 with the quality benchmark scores in turn, retain the captured data with scores greater than or equal to the quality benchmark scores, and summarize them to obtain the daily effective captured data of a single smart camera; S33: summing up the daily effective capture quantities of individual smart cameras within the analysis period to obtain the overall effective capture quantity; S34: Divide the total effective capture amount by the total capture amount in the analysis time period to obtain the total effective capture rate of a single smart camera.

[0011] In a further embodiment of the present invention, in step S4, the target recognition rate of the single intelligent camera is calculated as follows: S41: Analyze the target recognition status of each effective captured data of a single smart camera within the analysis time period one by one; S42: if there is no target recognition, mark the captured data; if there is target recognition but the target key business attribute is missing, mark the captured data; if there is target recognition but the target key business attribute information is incomplete, mark the captured data; if there is target recognition but there is manual feedback that the target recognition information is inaccurate, associate the captured data and mark it; S43: filtering the marked data in step S42 to obtain the number of target recognitions of a single smart camera within the analysis time period; S44: Divide the number of target recognitions by the total effective capture volume of the smart camera to obtain its target recognition rate.

[0012] In a further embodiment of the present invention: in step S5, the business results data is: The control and early warning data is generated through the data captured by a smart camera.

[0013] In a further embodiment of the present invention, in step S5, the calculation method of the success rate of the single smart camera is: S51: within the analysis period, the battle result data of each smart camera is counted; S52: Count the sum of the battle result data of the designated smart camera, and record it as the total battle result number; S53: Divide the battle result data of each smart camera by the total number of battle results to obtain the battle result generation rate of a single smart camera.

[0014] In a further solution of the present invention: in step S6, the availability index of the smart camera is established in the following manner: S611: In an analysis period, the four data elements of the online rate, the overall effective capture rate, the target recognition rate, and the battle result generation rate of a single smart camera are summed, and the value of each data element is compared with the sum of the data elements to obtain the probability of the corresponding data element; S612: In a plurality of consecutive analysis periods, the probabilities corresponding to the four data elements are calculated in sequence, thereby forming a probability array corresponding to each data element; S613: Calculate the entropy value of the probability array corresponding to each data element respectively; S614: Convert the entropy values ​​into difference coefficients respectively, and normalize the difference coefficients to obtain weights corresponding to the four data elements; S615: Multiply the values ​​corresponding to the four data elements in the current analysis period by the corresponding weights respectively, so as to obtain the current availability index of the smart camera.

[0015] In a further solution of the present invention: in step S6, the screening method of the availability of the smart camera is: S621: Calculate the mean and standard deviation of the availability of all smart cameras within an analysis period; S622: using the mean and standard deviation to set the first-level threshold, the second-level threshold, and the third-level threshold of the availability; S623: When the availability rate of the smart camera is less than the first threshold and greater than or equal to the second threshold, a mild unavailability warning is issued; when the availability rate of the smart camera is less than the second threshold and greater than or equal to the third threshold, a moderate unavailability warning is issued; when the availability rate of the smart camera is less than the third threshold, a severe unavailability warning is issued.

[0016] In a further embodiment of the present invention, in step S7, the smart camera early warning points system is: S711: Different scores are assigned to mild, moderate and severe warnings; S712: When the smart camera generates a corresponding warning, the corresponding score is automatically accumulated; S713: When the smart camera has only mild warnings, and the cumulative number does not exceed two, and no warning is generated within two consecutive analysis time periods after the latest warning, the warning score is cleared; when the smart camera has only moderate warnings, and the cumulative number does not exceed two, and no warning is generated within three consecutive analysis time periods after the latest warning, the warning score is cleared; when the smart camera has only one severe warning, and no warning is generated within four consecutive analysis time periods thereafter, the warning score is cleared; when the smart camera has only mild and moderate warnings, and the cumulative number does not exceed three, and no warning is generated within five consecutive analysis time periods after the latest warning, the warning score is cleared.

[0017] In a further solution of the present invention: in step S7, the clustering display mode of the integrating cameras is: S721: Filter smart cameras whose warning scores are not zero; S722: Sort by points from high to low; S723: Select smart cameras with large integrals in turn, and calculate all smart cameras with Euclidean distances within a threshold range to form a cluster; S724: Each cluster calculates the center point of its location for displaying the location of the cluster.

[0018] In a second aspect, the present invention further discloses a system for determining availability of the intelligent camera as described above, comprising: The data collection module is used to collect and classify the offline data of smart cameras, captured data, target recognition data, business results data, and manually marked data in the business system; A strategy configuration module is used to configure the strategy of calculating different types of data; Data calculation module, used to calculate different types of data according to the requirements of policy configuration; A data analysis module is used to establish an availability analysis model for the data results calculated by the data calculation module, and to perform fusion analysis on various types of data. Based on the analysis results, the smart cameras whose availability does not meet the requirements are screened, and a warning points system is established to convert availability warnings into points. The early warning management module is used to display early warnings and uniformly manage smart cameras whose availability does not meet the requirements; The visualization display module is used to realize cluster display of integrated cameras based on electronic maps, and to realize zone and block display according to the integrated situation and geographical location.

[0019] Beneficial effects of the present invention: (1) The present invention conducts availability evaluation for the current mainstream smart cameras, which meets the current and future development needs of video technology. It is also compatible with the evaluation of existing ordinary high-definition cameras (non-smart cameras). By setting the non-existent data items to zero, the dynamic adjustment of the availability index is achieved.

[0020] (2) The present invention abandons the previous method of evaluating camera availability through a single dimension. It comprehensively considers the online status, capture status, recognition status and application status of smart cameras in actual business, and conducts a comprehensive analysis. The availability of the camera is determined through the comprehensive analysis results. The analysis dimensions are more diversified and the results are more in line with reality.

[0021] (3) The present invention constructs a camera availability analysis model and adopts a method of automatic generation and dynamic change of weights to automatically analyze and generate sample data collected at different time periods, which meets the needs of current actual operation and avoids the adverse effects of manual assignment.

[0022] (4) The present invention realizes the dynamic division of camera availability, screens out smart cameras whose availability does not meet the requirements for graded warnings, and fully considers the actual operation situation, converts the warnings into integrals, and forms integral cameras, which have stronger data stability. The integral cameras are visualized in a clustering manner, so that their specific distribution can be viewed intuitively, which is convenient for guiding the maintenance of camera resources and further improving the overall availability of video resources.

[0023] (5) The present invention solves the problem of automatic identification of the availability of intelligent cameras. By defining four types of data elements and establishing an availability analysis model, the automatic calculation of camera availability indicators is realized. Through dynamic division of availability and early warning classification, the effective transformation of early warnings into points is realized. The actual situation of camera operation is fully considered, and the dynamic adjustment of points is realized. The entire process from data collection, classification processing, calculation analysis, and integral evaluation is completed, and a complete set of intelligent camera availability determination methods is built. The relationship and value between data are fully explored, and the intelligent screening of low-availability cameras is systematically realized. The needs of management and maintenance of various video surveillance systems are met, the operation and maintenance efficiency is greatly improved, and the utilization value of the entire video resource is enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present invention will be further described below in conjunction with the accompanying drawings.

[0025] Figure 1 It is an overall flow chart of a method for determining availability of an intelligent camera of the present invention; Figure 2 A flowchart of online rate calculation of a single smart camera of the present invention; Figure 3 A flowchart of calculating the overall effective capture rate of a single smart camera of the present invention; Figure 4 A flow chart showing the calculation of target recognition rate of a single intelligent camera of the present invention; Figure 5 A flowchart of calculating the success rate of a single smart camera of the present invention; Figure 6 A flow chart of a method for establishing an availability index of an intelligent camera of the present invention; Figure 7 It is a flow chart of the method for screening the availability of intelligent cameras of the present invention; Figure 8 It is a flow chart of the method for calculating the early warning score of the intelligent camera of the present invention; Fig. 9 A flow chart of the integrated camera clustering display method of the present invention; Fig.10 This is a diagram of the architecture of the intelligent camera availability determination system of the present invention; Fig.11 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] Example 1

[0028] See also Figure 1 As shown, the present invention is a method for determining the availability of an intelligent camera, which sequentially performs the following steps: Step 1: Obtain the online and offline status of the specified smart camera during the analysis period, and calculate the online rate of a single smart camera during the analysis period.

[0029] in: The method of obtaining the online and offline status data of the smart camera is as follows: Smart cameras are registered on the designated video management platform through a standard interface. Each camera is assigned a unique ID and maintains communication with the platform. The platform obtains the operating status of the connected smart camera by sending commands. If there is feedback, it means the camera is online. If there is no feedback, it means the camera is offline. The platform records the operating status and monitoring time of the smart camera and forms corresponding records. Through interface docking, all recorded data can be obtained to establish a camera status operation library.

[0030] The online rate of a single smart camera is calculated as follows: During the analysis period In the process, all corresponding records in the camera status runtime library are filtered according to the smart camera ID, and offline records are filtered according to the camera status field, as well as an online record adjacent to each offline record, and the same identifier is assigned; The records are sorted by time, and the offline records with the same identifier are extracted in sequence. The time of the offline record is subtracted from the time of the online record to obtain the single offline duration, and it is converted into hours. All single offline durations of a single smart camera are added together to obtain its offline duration in the analysis period. , is the number of smart cameras; Convert the total duration of the analysis period into hours to get the expected online time , subtract the offline time from the expected online time to get the actual online time of the smart camera during the analysis period ;

[0031] Actual online time Online time By comparison, the online rate of a single smart camera can be obtained. .

[0032]

[0033] Step 2: Count the amount of data captured by a single smart camera every day.

[0034] in: The captured data of the intelligent camera is obtained in real time from the video image analysis platform through the interface, including the camera number, camera name, target ID, capture time, image URL, attribute information, and quality score. The business attributes of different target objects are different. For example, for personnel targets, the business attribute information includes age, gender, whether they wear glasses, whether they wear masks, and whether they wear hats. For vehicle targets, the attribute information includes license plate number, vehicle color, and vehicle model. However, regardless of whether it is the captured data of personnel targets or vehicle targets, the quality score is provided by the corresponding algorithm.

[0035] The statistical method for calculating the daily capture volume of a single smart camera is as follows: For the captured data obtained by the interface, filter all the corresponding captured data by camera number; Further filtering is performed based on the capture time, and all the capture data of the smart camera within one day is retained; Filter duplicate captured data according to target ID and remove redundant captured data; Identify captured data where the image URL is unavailable and perform data filtering; Extract the time on the captured image and compare it with the recorded capture time. If the time difference is within the allowed time range, the data is retained; if the time difference exceeds the allowed time range, the data is filtered. After excluding the above filtered data, the daily capture volume of a single smart camera is counted .

[0036] Step 3: Evaluate the capture quality of the captured data, count the effective capture volume of a single smart camera per day and the overall effective capture volume during the analysis period, and calculate the overall effective capture rate of a single smart camera.

[0037] in: The overall effective capture rate of a single smart camera is calculated as follows: Set the image quality benchmark score; The quality scores of the daily snapshot data finally obtained in the second step are compared with the quality benchmark scores in turn, and the snapshot data with scores greater than or equal to the quality benchmark scores are retained and summarized to obtain the daily effective snapshot volume of a single smart camera. ; The analysis period The daily effective capture volume of each smart camera is summed up to get the overall effective capture volume ; Sum the daily snapshots of a single smart camera during the analysis period to get its total snapshot volume ; The total effective capture volume And the total number of snapshots during the analysis period Compared with the above, the overall effective capture rate of a single smart camera is obtained. .

[0038] .

[0039] Step 4: Analyze the target recognition status of the effective captured data to form the target recognition rate of a single smart camera within the analysis time period.

[0040] in: The target recognition rate of a single smart camera is calculated as follows: For a single smart camera in the analysis time period Analyze the target recognition status one by one based on the effective captured data; For those without target recognition, the captured data is marked; For targets that have been identified but whose key business attributes are missing, the captured data is marked; For targets that have been identified but have incomplete information on key business attributes, the captured data will be marked; For targets with identification, if there is manual feedback that the target identification information is inaccurate, the captured data is associated and marked; Filter the above labeled data to obtain the number of targets recognized by a single smart camera during the analysis period ; Identify the number of targets The total effective capture volume of the smart camera during the analysis period By comparison, the target recognition rate is obtained .

[0041] .

[0042] Step 5: Count the business performance data of the specified smart camera to form the performance rate of a single smart camera within the analysis period.

[0043] in: The business performance data of smart cameras refers to: Control and early warning data generated by capturing data from a smart camera; The calculation method of the success rate of a single smart camera is as follows: During the analysis period Statistics of each smart camera’s performance data ; The sum of the battle result data of the specified smart camera is counted and recorded as the total battle result number. ; The results data of each smart camera The total number of results By comparison, the success rate of a single smart camera is obtained. .

[0044] .

[0045] Step 6: Establish an availability index for smart cameras and screen out smart cameras whose availability does not meet the requirements for early warning.

[0046] in: The method for establishing the availability index of smart cameras is as follows: In an analysis period The online rate of a single smart camera , Overall effective capture rate , target recognition rate , Result Generation Rate Sum of four data elements ;

[0047] Set the camera online rate The sum of the data elements By comparison, we can get the probability of occurrence corresponding to the online rate of a single smart camera. ;

[0048] The overall effective capture rate The sum of the data elements By comparison, we can get the probability of occurrence corresponding to the overall effective capture rate of a single smart camera. ;

[0049] The target recognition rate The sum of the data elements By comparison, the probability of occurrence corresponding to the target recognition rate of a single smart camera is obtained. ;

[0050] The result rate The sum of the data elements By comparison, we can get the probability of occurrence corresponding to the success rate of a single smart camera. ;

[0051] In continuous Analysis period In the example, the probabilities corresponding to the four data elements are calculated in turn, thus forming a probability array corresponding to each data element. , , , ; The entropy value of the probability array corresponding to each data element is calculated respectively, where the entropy value corresponding to the online rate of a single smart camera is ;

[0052] The entropy value corresponding to the overall effective capture rate of a single smart camera is ;

[0053] The entropy value corresponding to the target recognition rate of a single smart camera is ;

[0054] The entropy value corresponding to the recognition rate of a single smart camera is ;

[0055] The above entropy values ​​are converted into difference coefficients respectively, where the difference coefficient corresponding to the online rate of a single smart camera is ;

[0056] The coefficient of difference corresponding to the overall effective capture rate of a single smart camera is ;

[0057] The coefficient of difference corresponding to the target recognition rate of a single smart camera is ;

[0058] The coefficient of difference corresponding to the recognition rate of a single smart camera is ;

[0059] The difference coefficients are normalized to obtain the weights corresponding to the four data elements. The weight corresponding to the online rate of a single smart camera is ;

[0060] The weight corresponding to the overall effective capture rate of a single smart camera is ;

[0061] The weight corresponding to the target recognition rate of a single smart camera is ;

[0062] The weight corresponding to the recognition rate of a single smart camera is ;

[0063] Multiply the four data elements in the current analysis period by the corresponding weights to obtain the current availability index of the smart camera. .

[0064]

[0065] The screening method for the availability of smart cameras is as follows: In an analysis period Calculate the average availability of all smart cameras and standard deviation ; Use the mean and standard deviation to set the first-level threshold of the availability of the smart camera , Secondary threshold , three-level threshold ; When the availability of smart cameras Less than the first threshold and greater than or equal to the second threshold, that is , to issue a mild unavailability warning; When the availability of smart cameras Less than the secondary threshold and greater than or equal to the tertiary threshold, that is , conduct moderate unavailability warning; When the availability of smart cameras Less than the third level threshold, that is , and issue a severe unavailability warning.

[0066] Step 7: Set different scores for different levels of warnings, build a smart camera warning points system, and realize clustering display of points cameras.

[0067] in: The smart camera early warning points are as follows: Different scores are assigned to mild, moderate and severe warnings. The score corresponding to mild warning is , the score corresponding to the moderate warning is , the score corresponding to the severe warning is , the score is set independently according to the actual situation; When the smart camera generates a corresponding warning, the corresponding score is automatically accumulated to obtain the total warning score. ; When the smart camera has only a minor warning, and the cumulative number does not exceed two, two consecutive analysis time periods after the latest warning If no warning is generated within a period, the total warning score will be reset to zero. ; When the smart camera has only moderate warnings, and the cumulative number does not exceed two, three consecutive analysis time periods after the latest warning If no warning is generated within a period, the total warning score will be reset to zero. ; When the smart camera has only one severe warning and there are four consecutive analysis time periods If no warning is generated within a period, the total warning score will be reset to zero. ; When the smart camera has only mild and moderate warnings, and the cumulative number does not exceed three, five consecutive analysis time periods after the latest warning If no warning is generated within a period, the total warning score will be reset to zero. ; The integrated camera clustering display is as follows: Filter the smart cameras whose total warning score is not zero, that is, ; Sort by points from high to low; Select the smart camera with the largest score and set a unified distance threshold , calculate the Euclidean distance between it and other integral cameras in turn , if the distance Less than or equal to , then the camera is included in the camera cluster In the process, until all the integral cameras have been compared; Remove clusters The integral camera in the cluster is selected, and the other intelligent cameras are selected with the largest integral, and the above steps are followed to form a camera cluster. ; And so on, multiple camera clusters are formed. ; For each camera cluster, the median of all longitudes and latitudes is calculated based on the longitude and latitude distribution of the cameras therein, and the center point of the camera cluster is obtained through the median of longitude and latitude for location display of the cluster.

[0068] Example 2

[0069] See also Figure 2 This embodiment makes a more specific implementation of the calculation method of the online rate of a single smart camera based on the embodiment 1, and the specific steps are as follows: S201, all smart cameras are registered to the designated video management platform through a standard interface, each camera is assigned a unique ID, and maintains communication with the platform; S202, the platform sends a command to the connected smart camera to obtain its operating status; S203, determining whether there is feedback for the command, if there is feedback, it means that the camera is online; if there is no feedback, it means that the camera is offline; S204, the platform records the online status, offline status and corresponding monitoring time of the smart camera to form a corresponding record; S205, through interface docking, all recorded data can be obtained and a camera status operation library can be established; S206, filtering the analysis time period in the camera status operation library according to the smart camera ID All the corresponding records in S207, filtering offline records according to the camera status field, and an online record adjacent to each offline record, and assigning the same identifier; S208, sorting the records by time, sequentially extracting offline records with the same identifier, subtracting the time of the offline record from the time of the online record to obtain a single offline duration, and converting it into hours; S209, in the analysis time period By adding up all the single offline durations of a single smart camera, the offline duration can be obtained. , is the number of smart cameras; S210, convert the total duration of the analysis time period into hours to obtain the expected online duration ; S211, the online time should be Subtract offline time , the actual online time of the smart camera during the analysis period can be obtained ;

[0070] S212, actual online time Online time By comparison, the online rate of a single smart camera can be obtained. .

[0071] .

[0072] Example 3

[0073] See also Figure 3 Based on Example 1, this embodiment makes a more specific implementation of the calculation method of the overall effective capture rate of a single smart camera. The specific steps are as follows: S301, the captured data of the intelligent camera is obtained in real time from the video image analysis platform through the interface, including the camera number, camera name, target ID, capture time, image URL, attribute information, and quality score. For human targets, the attribute information includes age, gender, whether wearing glasses, whether wearing a mask, and whether wearing a hat; for vehicle targets, the attribute information includes license plate number, vehicle color, and vehicle model; regardless of whether it is the captured data of human targets or vehicle targets, its quality score is provided by the corresponding algorithm.

[0074] S302, filtering all corresponding captured data by camera number for the captured data acquired by the interface; S303, filtering according to the capture time, and retaining all the capture data of the smart camera within one day; S304, filtering duplicate captured data according to the target ID, and removing redundant captured data; S305, identifying the captured data where the image URL is unavailable and performing data filtering; S306, extracting the time on the captured image, comparing it with the recorded capture time, and determining whether the time difference is within an allowable range; S307, if the time difference is within the allowed time range, jump to step S08; if the time difference exceeds the allowed time range, the data is filtered; S308: After excluding the above filtered data, the daily capture volume of a single smart camera is counted. .

[0075] S309, setting a picture quality benchmark score; S310, comparing the quality scores of the daily captured data finally obtained in step S308 with the quality benchmark scores in sequence; S311, determining whether the benchmark score is greater than or equal to the quality benchmark score; S312: If the score is greater than or equal to the quality benchmark score, the captured data is accumulated and summed to obtain the effective daily capture volume of a single smart camera. ; If it is less than the quality benchmark score, jump to step S310; S313, analyzing the time period The effective daily capture volume of a single smart camera Sum up and get the total effective capture amount ; S314, analyzing the time period The daily capture volume of a single smart camera Sum up and get the total capture amount ; S315, the total effective capture volume And the total number of snapshots during the analysis period Compared with the above, the overall effective capture rate of a single smart camera is obtained. .

[0076] .

[0077] Example 4

[0078] See also Figure 4 This embodiment makes a more specific implementation of the method for calculating the target recognition rate of a single smart camera based on the embodiment 1, and the specific steps are as follows: S401, analyzing a single smart camera in an analysis time period Valid captured data within S402, determining whether the captured data has target recognition; S403, if no, jump to step S410; if yes, jump to step S404; S404, for a target that has been identified, determining whether a key business attribute of the target is missing; S405, if yes, jump to step S410; if no, jump to step S406; S406, for those with target key business attributes, determine whether the attributes are complete; S407, if no, jump to step S410; if yes, jump to step S408; S408, for targets with complete key business attributes, determining whether there is manual feedback that target identification information is inaccurate; S409, if yes, jump to step S410; if no, jump to step S411; S410, marking the data; S411, filter the marked data to obtain the single smart camera in the analysis time period Number of target recognitions within ; S412, the number of target recognition The smart camera is used during the analysis period The total effective capture volume within By comparison, the target recognition rate is obtained ; .

[0079] Example 5

[0080] See also Figure 5 Based on Example 1, this embodiment makes a more specific implementation of the calculation method of the battle result generation rate of a single smart camera. The specific steps are as follows: S501, in the analysis time period Statistics of each smart camera’s performance data ; S502: Count the sum of the battle result data of the designated smart camera and record it as the total battle result number. ; S503, the battle result data of each smart camera The total number of results By comparison, the success rate of a single smart camera is obtained. ; .

[0081] Example 6

[0082] See also Figure 6 This embodiment makes a more specific implementation of the method for establishing the availability index of the smart camera based on the embodiment 1, and the specific steps are as follows: S601, in an analysis period Within, the online rate of a single smart camera , Overall effective capture rate , target recognition rate , Result Generation Rate Sum of four data elements ; S602, the online rate of the cameras is adjusted in sequence. , Overall effective capture rate , target recognition rate , Result Generation Rate The sum of the data elements By comparison, we can obtain the occurrence probability corresponding to the online rate of a single smart camera. , the probability of occurrence corresponding to the overall effective capture rate , the probability of occurrence corresponding to the target recognition rate The probability of occurrence corresponding to the battle result generation rate ; S603, in continuous Analysis period In the example, the probabilities corresponding to the four data elements are calculated in turn, thus forming a probability array corresponding to each data element. , , , ; S604, respectively calculating the entropy value of the probability array corresponding to each data element, where the entropy value corresponding to the online rate of a single smart camera is ;

[0083] The entropy value corresponding to the overall effective capture rate of a single smart camera is ;

[0084] The entropy value corresponding to the target recognition rate of a single smart camera is ;

[0085] The entropy value corresponding to the recognition rate of a single smart camera is ;

[0086] S605, convert the above entropy values ​​into difference coefficients respectively, where the difference coefficient corresponding to the online rate of a single smart camera is ;

[0087] The coefficient of difference corresponding to the overall effective capture rate of a single smart camera is ;

[0088] The coefficient of difference corresponding to the target recognition rate of a single smart camera is ;

[0089] The coefficient of difference corresponding to the recognition rate of a single smart camera is ;

[0090] S606, normalize the difference coefficients to obtain weights corresponding to the four data elements, where the weight corresponding to the online rate of a single smart camera is ; ; The weight corresponding to the overall effective capture rate of a single smart camera is ; ; The weight corresponding to the target recognition rate of a single smart camera is ; ; The weight corresponding to the recognition rate of a single smart camera is ; ; S607, multiply the four data elements in the current analysis period by the corresponding weights to obtain the current availability index of the smart camera ; .

[0091] Example 7

[0092] See also Figure 7 This embodiment makes a more specific implementation of the screening method of the availability of smart cameras based on the embodiment 1, and the specific steps are as follows: S701, in an analysis period Calculate the average availability of all smart cameras and standard deviation ; S702, using the mean and standard deviation to set the first-level threshold of the availability of the smart camera , Secondary threshold , three-level threshold ; S703, comparing the availability of each smart camera with a threshold in turn; S704, determining the availability of smart cameras Is it less than the first level threshold? ; S705, if no, jump to step S703; if yes, jump to step S706; S706, determining the availability of the smart camera Is it greater than or equal to the secondary threshold? ; S707: If yes, issue a mild unavailability warning; if no, jump to step S708; S708, determining the availability of smart cameras Is it greater than or equal to the third level threshold? ; S709: If yes, perform a moderate unavailability warning; if no, jump to step S710; S710: Issue a severe unavailability warning.

[0093] Example 8

[0094] See also Figure 8 This embodiment makes a more specific implementation of the early warning score calculation method of the smart camera based on the embodiment 1, and the specific steps are as follows: S801, assign different scores to mild, moderate and severe warnings. The score corresponding to mild warning is , the score corresponding to the moderate warning is , the score corresponding to the severe warning is , the score is set independently according to the actual situation; S802, when the smart camera generates a corresponding warning, the corresponding score is automatically accumulated to obtain the total warning score ; S803, counting the number of warning types of a single smart camera; S804, determining whether the number of warning types is 3; S805, if yes, jump to step S802; if no, jump to step S806; S806, determining whether the number of warning types is 2; S807, if yes, jump to step S808; if no, jump to step S814; S808, determining whether there are only mild warnings and moderate warnings; S809, if yes, jump to step S810; if no, jump to step S802; S810, determining whether the cumulative number of warnings exceeds three times; S811, if so, jump to step S802; if not, jump to step S812; S812, determine if there are five consecutive analysis time periods after the most recent warning Is there any warning? S813, if yes, jump to step S830; if no, jump to step S802; S814, determining whether it is a mild warning; S815, if yes, jump to step S816; if no, jump to step S820; S816, determining whether the cumulative number of mild warnings exceeds two; S817, if yes, jump to step S802; if no, jump to step S818; S818, determine two consecutive analysis time periods after the most recent warning Is there any warning? S819, if yes, jump to step S830; if no, jump to step S802; S820, determining whether it is a moderate warning; S821, if yes, jump to step S822; if no, jump to step S826; S822, determining whether the cumulative number of moderate warnings exceeds two; S823, if so, jump to step S802; if not, jump to step S824; S824, determine if there are three consecutive analysis time periods after the most recent warning Is there any warning? S825, if yes, jump to step S830; if no, jump to step S802; S826, determining whether the number of severe warnings is one; S827, if yes, jump to step S828; if no, jump to step S802; S828, determine that there are four consecutive analysis time periods after the most recent warning Is there any warning? S829, if yes, jump to step S830; if no, jump to step S802; S830, the total warning score is cleared.

[0095] Example 9

[0096] See also Fig. 9 This embodiment makes a more specific implementation of the clustering display method of the integral camera based on the embodiment 1, and the specific steps are as follows: S901, screening smart cameras whose total warning score is not zero; S902, sorting the scores from large to small; S903, setting a unified distance threshold , select the smart camera with the largest integral, and calculate the Euclidean distance between it and other integral cameras in turn , if the distance Less than or equal to , then the camera is included in the camera cluster In the process, until all the integral cameras have been compared; S904, remove clustering The other intelligent cameras with the largest score are selected and the operation is performed according to step S903 to form a camera cluster. ; S905, and so on, until all smart cameras are classified, thereby forming multiple camera clusters. ; S906: For each camera cluster, according to the longitude and latitude distribution of the cameras therein, the median of all longitudes and latitudes is calculated respectively, and the center point of the camera cluster is obtained by the median of the longitude and latitude for displaying the location of the cluster.

[0097] Example 10

[0098] See also Fig.10 Based on Example 1, this embodiment discloses an intelligent camera availability determination system, including a data acquisition module, a policy configuration module, a data calculation module, a data analysis module, an early warning management module, and a visualization display module.

[0099] in: The data collection module collects offline data, captured data, target recognition data, and business results data from smart cameras. At the same time, it collects manually marked data in the business system and manages the data by classification. The strategy configuration module is used to support the data calculation module and realize the configuration of various calculation methods for different data; The data calculation module is used to calculate different types of data according to the requirements of policy configuration; The data analysis module uses the result data calculated by the data calculation module to establish an availability analysis model, realize the fusion analysis of various types of data, and screen out smart cameras whose availability does not meet the requirements based on the analysis results; and build an early warning points system to convert availability warnings into points to form point cameras; The early warning management module provides early warning display and unified management for smart cameras whose availability does not meet the requirements; The visualization display module realizes cluster display of integrated cameras based on electronic maps, and realizes zone and block display according to the integrated situation and geographical location.

[0100] In summary, the present invention solves the problem of automatic identification of the availability of intelligent cameras. By clarifying four types of data elements and establishing an availability analysis model, the automatic calculation of camera availability indicators is realized. Through dynamic division of availability and early warning classification, the effective conversion of early warning to integral is realized. The actual situation of camera operation is fully considered to realize dynamic adjustment of integral. The entire process from data collection, classification processing, calculation analysis, and integral evaluation is completed. A complete set of intelligent camera availability determination methods is built to realize intelligent screening of low-availability cameras, solve the singleness and limitations of traditional analysis methods, fully explore the relationship and value between data, give play to the advantages of fusion data analysis, realize dynamic analysis of availability through models, reduce manual intervention, and improve the adaptability of analysis methods. It meets the management needs of public security, urban management, gardening, cultural tourism and other departments as well as enterprises, institutions and communities to build their own video surveillance systems, greatly improves the problem discovery rate and overall operation and maintenance efficiency, and improves the utilization value of the entire video resource.

[0101] Embodiment 11

[0102] See also Fig.11 This embodiment provides an electronic device, including a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304.

[0103] in: The memory 303 is used to store computer programs; When the processor 301 is used to execute the program stored in the memory 303, the following steps are implemented: Collect offline data, captured data, target recognition data, business results data, and manually marked data from business systems from smart cameras, classify and manage the data, and provide a data foundation; Configure calculation methods for different data and perform calculations on different types of data; An availability analysis model is established for the calculated result data to realize the fusion analysis of various data. Based on the analysis results, smart cameras whose availability does not meet the requirements are screened out, and an early warning points system is established to convert availability warnings into points to form point cameras. Provide early warning display and unified management for smart cameras whose availability does not meet the requirements; Rely on electronic maps to realize cluster display of integration cameras, and realize zone and block presentation according to integration situation and geographical location.

[0104] The communication bus mentioned in the above electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0105] The communication interface is used for communication between the above electronic device and other devices.

[0106] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0107] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0108] Example 12

[0109] This embodiment provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the above-mentioned methods for determining the availability of an intelligent camera are implemented.

[0110] Example 13

[0111] This embodiment provides a computer program product including instructions, which, when executed on a computer, enables the computer to execute any one of the methods for determining the availability of an intelligent camera in the above embodiments.

[0112] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive Solid State Disk (SSD)), etc.

[0113] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0114] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system, electronic device and storage medium, since they are basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0115] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for determining the availability of an intelligent camera, characterized in that: Follow these steps: S1: Obtain the online and offline status of a specified smart camera during the analysis period, and calculate the online rate of a single smart camera during the analysis period; S2: count the number of captured data of a single smart camera every day; S3: Evaluate the capture quality of the captured data, count the effective capture volume of a single smart camera per day and the overall effective capture volume in the analysis period, and calculate the overall effective capture rate of a single smart camera; S4: Analyze the target recognition status of the effective captured data to form the target recognition rate of a single smart camera within the analysis time period; S5: Count the business performance data of the specified smart camera to form the performance rate of a single smart camera within the analysis period; S6: Establish an availability index for smart cameras, and screen out smart cameras whose availability does not meet the requirements for early warning; S7: Set different scores for different levels of warnings, build a smart camera warning points system, and realize cluster display of scoring cameras.

2. The method for determining the availability of an intelligent camera according to claim 1, characterized in that: In step S1, the online rate of a single smart camera in the analysis time period is calculated as follows: S11: count the offline time of a single smart camera; S12: converting the total duration of the analysis period into the expected online duration, and obtaining the actual online duration of the smart camera in the analysis period by the difference between the expected online duration and the offline duration; S13: Calculate the ratio of the actual online duration to the expected online duration to obtain the online rate of a single smart camera.

3. The method for determining the availability of an intelligent camera according to claim 1, characterized in that: In step S2, the captured data of the single smart camera includes the camera number, camera name, target ID, capture time, picture URL, attribute information, and quality score; The calculation method of the daily capture volume of a single smart camera is: S21: Taking a single smart camera as a unit, obtaining its captured data by camera number; S22: Filter and retain all captured data of the smart camera within one day according to the capture time; S23: Filter the repeated captured data according to the target ID, and perform data filtering; identify the captured data with unavailable image URLs, and perform data filtering; identify the captured data with large capture time deviations, and perform data filtering; S24: After excluding the filtered data in step S23, the daily capture volume of a single smart camera is counted.

4. The method for determining the availability of an intelligent camera according to claim 3, characterized in that: In step S3, the overall effective capture rate of the single smart camera is calculated as follows: S31: Setting the picture quality benchmark score; S32: Compare the quality scores of the daily captured data in step S2 with the quality benchmark scores in turn, retain the captured data with scores greater than or equal to the quality benchmark scores, and summarize them to obtain the daily effective captured data of a single smart camera; S33: summing up the daily effective capture quantities of individual smart cameras within the analysis period to obtain the overall effective capture quantity; S34: Divide the total effective capture amount by the total capture amount in the analysis time period to obtain the total effective capture rate of a single smart camera.

5. The method for determining the availability of an intelligent camera according to claim 1, characterized in that: In step S4, the target recognition rate of the single smart camera is calculated as follows: S41: Analyze the target recognition status of each effective captured data of a single smart camera within the analysis time period one by one; S42: if there is no target recognition, mark the captured data; if there is target recognition but the target key business attribute is missing, mark the captured data; if there is target recognition but the target key business attribute information is incomplete, mark the captured data; if there is target recognition but there is manual feedback that the target recognition information is inaccurate, associate the captured data and mark it; S43: filtering the marked data in step S42 to obtain the number of target recognitions of a single smart camera within the analysis time period; S44: Divide the number of target recognitions by the total effective capture volume of the smart camera to obtain its target recognition rate.

6. The method for determining the availability of an intelligent camera according to claim 1, characterized in that: In step S5, the business results data is: Control and early warning data generated by capturing data from a smart camera; The calculation method of the success rate of a single smart camera is: S51: within the analysis period, the battle result data of each smart camera is counted; S52: Count the sum of the battle result data of the designated smart camera, and record it as the total battle result number; S53: Divide the battle result data of each smart camera by the total number of battle results to obtain the battle result generation rate of a single smart camera.

7. The method for determining the availability of an intelligent camera according to claim 1, characterized in that: In step S6, the availability index of the smart camera is established as follows: S611: In an analysis period, the four data elements of the online rate, the overall effective capture rate, the target recognition rate, and the battle result generation rate of a single smart camera are summed, and the value of each data element is compared with the sum of the data elements to obtain the probability of the corresponding data element; S612: In a plurality of consecutive analysis periods, the probabilities corresponding to the four data elements are calculated in sequence, thereby forming a probability array corresponding to each data element; S613: Calculate the entropy value of the probability array corresponding to each data element respectively; S614: Convert the entropy values ​​into difference coefficients respectively, and normalize the difference coefficients to obtain weights corresponding to the four data elements; S615: Multiply the values ​​corresponding to the four data elements in the current analysis period by the corresponding weights respectively, so as to obtain the current availability index of the smart camera; The screening method of the availability of the smart camera is: S621: Calculate the mean and standard deviation of the availability of all smart cameras within an analysis period; S622: using the mean and standard deviation to set the first-level threshold, the second-level threshold, and the third-level threshold of the availability; S623: When the availability rate of the smart camera is less than the first threshold and greater than or equal to the second threshold, a mild unavailability warning is issued; when the availability rate of the smart camera is less than the second threshold and greater than or equal to the third threshold, a moderate unavailability warning is issued; when the availability rate of the smart camera is less than the third threshold, a severe unavailability warning is issued.

8. The method for determining the availability of an intelligent camera according to claim 1, characterized in that: In step S7, the smart camera early warning points system is: S711: Different scores are assigned to mild, moderate and severe warnings; S712: When the smart camera generates a corresponding warning, the corresponding score is automatically accumulated; S713: When the smart camera has only a minor warning, and the cumulative number does not exceed two, and no warning is generated within two consecutive analysis time periods after the latest warning, the warning score is cleared; When the smart camera has only one moderate warning, and the cumulative number does not exceed two, and no warning is generated within three consecutive analysis time periods after the latest warning, the warning score is cleared; when the smart camera has only one severe warning, and no warning is generated within four consecutive analysis time periods, the warning score is cleared; When the smart camera has only mild and moderate warnings, and the cumulative number does not exceed three, and no warnings are generated within five consecutive analysis time periods after the latest warning, the warning score will be cleared.

9. The method for determining the availability of an intelligent camera according to claim 11, characterized in that: In step S7, the clustering display mode of the integration cameras is: S721: Filter smart cameras whose warning scores are not zero; S722: Sort by points from high to low; S723: Select smart cameras with large scores in turn, and calculate all smart cameras with Euclidean distances within a threshold range to form a cluster; S724: Each cluster calculates the center point of its location for displaying the location of the cluster.

10. A smart camera availability determination system based on any one of claims 1-9, characterized in that: include: The data collection module is used to collect and classify the offline data of smart cameras, captured data, target recognition data, business results data, and manually marked data in the business system; A strategy configuration module is used to configure the strategy of calculating different types of data; Data calculation module, used to calculate different types of data according to the requirements of policy configuration; A data analysis module is used to establish an availability analysis model for the data results calculated by the data calculation module, and to perform fusion analysis on various types of data. Based on the analysis results, the smart cameras whose availability does not meet the requirements are screened, and a warning points system is established to convert availability warnings into points. The early warning management module is used to display early warnings and uniformly manage smart cameras whose availability does not meet the requirements; The visualization display module is used to realize cluster display of integrated cameras based on electronic maps, and to realize zone and block display according to the integrated situation and geographical location.

Images