Abstract: A method of calculating a size of a Gaussian kernel in a crowd counting system. The method includes dividing a crowd density map into several areas at equal intervals; calculating an estimated value of average distance of k nearest heads with respect to a head center point coordinate of a current person; calculating a weight coefficient for an average distance of k nearest heads; calculating a size of a Gaussian kernel corresponding to the head center point coordinate of the current person; determining whether there are any head center point coordinates of which sizes of Gaussian kernels have not been calculated yet in training data. If there are head center point coordinates of which sizes of the Gaussian kernels have not been calculated in the training data, recalculating an estimated value of average distance of k nearest heads with respect to another head center point coordinate of another current person.

Abstract: Provided are a road safety assessment method and apparatus, and an electronic device and a readable medium. The method comprises: acquiring real-time water level data in a culvert and real-time roadbed state data of a road within a preset distance range from the culvert (S201); on the basis of the roadbed state data, determining anti-pressure capability data of the road (S202); and on the basis of the water level data and the anti-pressure capability data, assessing the safety of the road to obtain a road safety assessment result (S203). The accuracy of road safety assessment is improved, and unattended monitoring without interruption can be realized.

Abstract: The present disclosure relates to the technical field of video monitor information. Provided are a video patrol method and apparatus, an electronic device, and a readable medium. The video patrol method comprises: obtaining video patrol information, wherein the video patrol information is video information obtained from a real scene; matching and combining the video patrol information with a 3D roaming scenario to obtain a 3D patrol video, wherein the 3D roaming scenario is a virtual scenario obtained by simulating the real scene; and displaying the 3D patrol video. The present disclosure can reduce the labor intensity of operation and maintenance workers, and improve the patrol efficiency.

Abstract: The present disclosure provides a fault location method, a fault location system, and a video management system, and relates to the technical field of network communication. The fault location method includes: determining, in response to a case where it is determined that a target video from a video acquisition system has a fault, a network identifier of a target camera corresponding to the target video; extracting target alarm data corresponding to the network identifier from current alarm data of the video acquisition system according to the network identifier of the target camera; and determining a fault cause of the video acquisition system according to a preset corresponding relationship between fault events and alarm information, and the target alarm data. The present disclosure can be used for centralized management and operation and maintenance of devices and systems related to video services.

Abstract: A method and system for uniformly processing logs of multiple formats under a security situation awareness system. The method includes defining a universal interface file and an interface file that corresponds to each device ID of each vendor; collecting log files of respective vendors; putting a file transfer protocol into the collected log files and the defined universal interface file; reading, when change of any log file is monitored, the log file line by line, and updating the log file through the file transfer protocol; identifying a corresponding device ID; screening out an interface file corresponding to the device ID; based on the screened interface file, converting the updated log file into an interpretable uniform format in terms of the universal interface file; and displaying graphically a log file resulted from the uniform format, and completing a uniform processing with respect to the logs of multiple formats.

Abstract: A method of calculating a size of a Gaussian kernel in a crowd counting system. The method includes dividing a crowd density map into several areas at equal intervals; calculating an estimated value of average distance of k nearest heads with respect to a head center point coordinate of a current person; calculating a weight coefficient for an average distance of k nearest heads; calculating a size of a Gaussian kernel corresponding to the head center point coordinate of the current person; determining whether there are any head center point coordinates of which sizes of Gaussian kernels have not been calculated yet in training data. If there are head center point coordinates of which sizes of the Gaussian kernels have not been calculated in the training data, recalculating an estimated value of average distance of k nearest heads with respect to another head center point coordinate of another current person.

Abstract: Our invention CHARM for preventing children from being left behind in a car has the potential to dramatically reduce the number of heat stroke deaths. The invention leverages emerging high technologies (such as sensors for accelerometer and temperature, Arduino microcontroller) and a circuit board in connection with car monitor and control system (such as panic alarm and door system) to detect if a child is left behind in a car and send out an alarm.