CN-122024165-A - Visual camp monitoring method and system based on intelligent guard

CN122024165ACN 122024165 ACN122024165 ACN 122024165ACN-122024165-A

Abstract

Visual camp monitoring method and system based on intelligent guard and guard relates to the camp management field. The method comprises the steps of obtaining multiple environmental data and image data of each sentinel in a camp, carrying out quality evaluation on the image data to obtain a corresponding image quality change curve, summarizing the curve to construct an image quality change band, calculating correlation coefficients of the environmental data and the change band width curve to determine influence coefficients of the environmental data, inputting the environmental data into a preset prediction model to obtain camp environmental prediction data of the data, predicting subsequent image quality change of each sentinel by combining the influence coefficients, the environmental prediction data and the quality change curve, calculating a comprehensive evaluation value of the subsequent image quality of each sentinel to determine a monitoring operation adjustment priority of each sentinel, and finally carrying out dynamic adjustment on the subsequent visual monitoring strategy of each sentinel according to the priority, so that reasonable allocation of camp monitoring resources is realized, and visual monitoring efficiency of the camp is improved.

Inventors

WANG JUN
LIU XIONG
ZHU QIAN
LIU BENJUN
XIANG LIANG
QUAN JIANG

Assignees

湖北集防科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260209

Claims (9)

1. The visual camp monitoring method based on the intelligent guard is characterized by being applied to a barracks security control platform, and comprises the following steps: Acquiring various environmental data of the camp in a preset first time period and image data uploaded by a plurality of intelligent on-duty sentries; performing image quality evaluation on the image data uploaded by each intelligent on-duty guard in the preset first time period to obtain an image quality change curve corresponding to each intelligent on-duty guard; Summarizing image quality change curves corresponding to all intelligent on-duty guard posts, and constructing an image quality change band, wherein the bandwidth of the image quality change band represents the difference between the maximum value and the minimum value of the image quality evaluation values of all intelligent on-duty guard posts at the same acquisition time; Calculating a correlation coefficient between each piece of environment data and the bandwidth curve of the image quality change band to obtain an influence coefficient of each piece of environment data; inputting a plurality of environmental data into a preset environmental prediction model to obtain camping environment prediction data corresponding to the environmental data in a preset second time period; Generating an image quality change prediction curve of each intelligent duty guard in the preset second time period based on the influence coefficient of each environmental data, the camp environment prediction data and the image quality change curve corresponding to each intelligent duty guard; Calculating an image quality comprehensive evaluation value of an image quality change prediction curve of each intelligent on-duty guard, and determining a monitoring operation adjustment priority of each intelligent on-duty guard in the preset second time period according to the image quality comprehensive evaluation value; and adjusting the visual monitoring strategy of each intelligent on-duty guard within the preset second time period based on the monitoring operation adjustment priority of each intelligent on-duty guard.
2. The method according to claim 1, wherein the image quality evaluation is performed on the image data uploaded by each intelligent on-duty guard in the preset first period of time to obtain an image quality change curve corresponding to each intelligent on-duty guard, specifically: Inputting an image to be evaluated uploaded by a first intelligent on-duty whistle at any time point in the preset first time period into a preset target recognition model to obtain a recognizable area and an unrecognizable area of the image to be evaluated, wherein the first intelligent on-duty whistle is any one of a plurality of intelligent on-duty whistles; Calculating the ratio of the total pixel area of the identifiable region to the total pixel area of the image to be evaluated to obtain the identifiable region duty ratio of the image to be evaluated; Extracting pixel coordinates of each pixel point in the identifiable region, and calculating an average value of Euclidean distances between every two pixel coordinates to obtain the region dispersion of the identifiable region; Calculating an image quality evaluation value of the image to be evaluated based on the identifiable region duty ratio of the image to be evaluated and the region dispersion of the identifiable region; Summarizing the image quality evaluation values of the images to be evaluated uploaded by the first intelligent on-duty guard at all time points in the preset first time period, and constructing an image quality change curve corresponding to the first intelligent on-duty guard by taking time as a horizontal axis and the image quality evaluation value as a vertical axis.
3. The method according to claim 1, wherein the calculating a correlation coefficient between each of the environmental data and the bandwidth curve of the image quality variation band obtains an influence coefficient of each of the environmental data, specifically: calculating the fluctuation amplitude of each bandwidth in the bandwidth curve of the image quality variation band; dividing a bandwidth curve of the image quality variation band into a stable bandwidth curve and a abrupt bandwidth curve based on a preset image quality fluctuation amplitude threshold; Calculating a first correlation coefficient between the first environmental data and the stable bandwidth curve by adopting a Pierson correlation coefficient method, and calculating a second correlation coefficient between the first environmental data and the abrupt bandwidth curve by adopting a Stersman rank correlation coefficient method, wherein the first environmental data is any one of a plurality of environmental data; and carrying out weighted fusion on the first correlation coefficient and the second correlation coefficient to obtain an influence coefficient of the first environmental data on image quality.
4. The method of claim 1, wherein the predetermined environmental prediction model comprises an environmental factor coupling module and an environmental data prediction module, the environmental factor coupling module is a multi-factor coupling matrix constructed by correlation coefficients between a plurality of environmental data in the predetermined first period, and the environmental data prediction module is an LSTM time series prediction network trained from the barrage historical time series environmental data.
5. The method of claim 4, wherein the multi-factor coupling matrix is constructed in the following manner: Calculating correlation coefficients between the plurality of environmental data; The correlation coefficients between the plurality of environmental data are arranged in a matrix mode according to the attribute of the environmental data, and an initial multi-factor coupling matrix is obtained; And taking 0 from elements in the initial multi-factor coupling matrix, which are positioned in the preset association strength threshold range, based on the preset association strength threshold range to obtain the multi-factor coupling matrix.
6. The method according to claim 1, wherein the generating an image quality change prediction curve of each intelligent guard over the preset second period based on the influence coefficient of each environmental data, the camp environment prediction data, and the image quality change curve corresponding to each intelligent guard, specifically comprises: a. Extracting an image quality evaluation value of a second intelligent on-duty whistle at a first time point from an image quality change curve corresponding to the second intelligent on-duty whistle, wherein the second intelligent on-duty whistle is any one of a plurality of intelligent on-duty whistles, and the first time point is the last time point in the image quality change curve of the second intelligent on-duty whistle; b. Extracting predicted environmental data at a second time point from environmental predicted data of a camp environment of second environmental data, and extracting real-time environmental data at the first time point from the environmental data uploaded by the second intelligent guard post, wherein the second environmental data is any one of a plurality of environmental data, and the second time point is the first time point after the first time point; c. calculating to obtain the environment deviation degree of the second time point according to the predicted environment data of the second time point and the real-time environment data of the first time point; d. Calculating to obtain an environment deviation value of the second time point according to the influence coefficient of the second environment data and the environment deviation degree of the second time point; e. Calculating to obtain an image quality deviation value of the second environment data, which is caused by uploading image data by the second intelligent on-duty guard at the second time point, based on the environment deviation value of the second time point and the image quality evaluation value of the second intelligent on-duty guard at the first time point; f. summing the image quality deviation values caused by the image data uploaded by the second intelligent on-duty guard at the second time point by the environment data to obtain the image quality integral deviation value of the image data uploaded by the second intelligent on-duty guard at the second time point; g. Summing the integral deviation value of the image quality of the image data uploaded by the second intelligent on-duty guard at the second time point with the image quality evaluation value of the image data uploaded by the second intelligent on-duty guard at the first time point to obtain an image quality prediction evaluation value of the image data uploaded by the second intelligent on-duty guard at the second time point; h, taking the second time point as the first time point; repeating the steps a to h to obtain image quality prediction evaluation values of the second intelligent guard at all time points within the preset second time period, and constructing an image quality change prediction curve of the second intelligent duty guard.
7. Visual monitoring system in camp based on intelligence guard on duty sentry, its characterized in that, the system is the district security protection control platform, the system includes acquisition module (1), processing module (2) and control module (3), wherein: the acquisition module (1) is used for acquiring various environmental data and image data uploaded by a plurality of intelligent on-duty sentry on the camp within a preset first time period; The processing module (2) is used for carrying out image quality evaluation on the image data uploaded by each intelligent on-duty guard in the preset first time period to obtain an image quality change curve corresponding to each intelligent on-duty guard; Summarizing image quality change curves corresponding to all intelligent on-duty guard posts, and constructing an image quality change band, wherein the bandwidth of the image quality change band represents the difference between the maximum value and the minimum value of the image quality evaluation values of all intelligent on-duty guard posts at the same acquisition time; Calculating a correlation coefficient between each piece of environment data and the bandwidth curve of the image quality change band to obtain an influence coefficient of each piece of environment data; inputting a plurality of environmental data into a preset environmental prediction model to obtain camping environment prediction data corresponding to the environmental data in a preset second time period; Generating an image quality change prediction curve of each intelligent duty guard in the preset second time period based on the influence coefficient of each environmental data, the camp environment prediction data and the image quality change curve corresponding to each intelligent duty guard; Calculating an image quality comprehensive evaluation value of an image quality change prediction curve of each intelligent on-duty guard, and determining a monitoring operation adjustment priority of each intelligent on-duty guard in the preset second time period according to the image quality comprehensive evaluation value; The control module (3) is used for adjusting the priority based on the monitoring operation of each intelligent on-duty guard, and adjusting the visual monitoring strategy of each intelligent on-duty guard in the preset second time period.
8. An electronic device comprising a processor (301), a memory (305), a user interface (303) and a network interface (304), the memory (305) being for storing instructions, the user interface (303) and the network interface (304) being for communicating to other devices, the processor (301) being for executing the instructions stored in the memory (305) to cause the electronic device (300) to perform the method according to any one of claims 1 to 6.
9. A computer readable storage medium storing instructions which, when executed, perform the method of any one of claims 1 to 6.

Description

Visual camp monitoring method and system based on intelligent guard Technical Field The application relates to the field of camp management, in particular to a camp visual monitoring method and system based on intelligent guard whistle. Background The camp is used as an important scene for regional safety and emergency guarantee, and the guarantee of the safety of the camp has very important practical significance for regional stability. In traditional camp security, camp security mainly relies on a fixedly deployed monitoring camera, manual patrol and a physical barrier, but the mode has the problems of large monitoring blind area, low response speed, high labor cost and the like. Along with the development of the Internet of things, artificial intelligence and robot technology, an intelligent security system is gradually applied to camp management, the intelligent security system is formed by a monitoring platform and a plurality of intelligent duty guard posts in a cooperative mode, wherein the intelligent duty guard posts comprise a plurality of forms such as a fixed intelligent monitoring tower, a ground mobile robot and an aerial duty guard unmanned plane, various sensors such as a high-definition camera, an infrared thermal imager and a radar are integrated, each intelligent duty guard post is distributed to a preset patrol route or a fixed guard point, then a remote instruction is autonomously or accepted, the tasks such as regional patrol, image acquisition, target identification and alarm are executed, and the monitoring platform is responsible for gathering monitoring data from all intelligent duty guard posts, and centralized display, comprehensive analysis and alarm processing are carried out, so that the intelligent security level of the camp is comprehensively improved. However, the actual campsite scene is often located in complex environments such as mountain areas and woodlands, the visual monitoring environment in the campsite is in a fluctuation state for a long time, at this time, in order to ensure the visual monitoring precision of each intelligent on-duty whistle, the monitoring strategy of each intelligent on-duty whistle needs to be adjusted based on the environment fluctuation, for example, the acquisition frequency is improved when the environment fluctuation is large, but due to the huge campsite range, the number of intelligent on-duty whistles is numerous, when the visual monitoring environment fluctuation is severe, the system can have the problems of strategy calculation delay and scheduling instruction congestion due to limited calculation resources, so that the monitoring strategy cannot fall to the ground in time, and the visual monitoring efficiency is greatly limited. Disclosure of Invention Aiming at the problem that the visual monitoring efficiency is reduced because intelligent on-duty sentry monitoring is difficult to land in time in the existing camp, the application provides a camp visual monitoring method and system based on the intelligent on-duty sentry. In a first aspect, the application provides a camp visual monitoring method based on an intelligent guard post, which is applied to a barracking security control platform, and comprises the following steps: Acquiring various environmental data of the camp in a preset first time period and image data uploaded by a plurality of intelligent on-duty sentries; performing image quality evaluation on the image data uploaded by each intelligent on-duty guard in the preset first time period to obtain an image quality change curve corresponding to each intelligent on-duty guard; Summarizing image quality change curves corresponding to all intelligent on-duty guard posts, and constructing an image quality change band, wherein the bandwidth of the image quality change band represents the difference between the maximum value and the minimum value of the image quality evaluation values of all intelligent on-duty guard posts at the same acquisition time; Calculating a correlation coefficient between each piece of environment data and the bandwidth curve of the image quality change band to obtain an influence coefficient of each piece of environment data; inputting a plurality of environmental data into a preset environmental prediction model to obtain camping environment prediction data corresponding to the environmental data in a preset second time period; Generating an image quality change prediction curve of each intelligent duty guard in the preset second time period based on the influence coefficient of each environmental data, the camp environment prediction data and the image quality change curve corresponding to each intelligent duty guard; Calculating an image quality comprehensive evaluation value of an image quality change prediction curve of each intelligent on-duty guard, and determining a monitoring operation adjustment priority of each intelligent on-duty guard in the preset second time period according to the