Real-time video analysis system and method based on b/s architecture

Abstract

The present invention provides a real-time video analysis system based on B / S architecture. The Web front-end module receives user input instructions, sends the user input instructions to the scheduling server module, retrieves the analysis results from the scheduling server module, and presents the analysis results; the scheduling server The module starts the main analysis module according to the user input instruction, sends the user input instruction to the instruction forwarding module and receives the analysis result and temporarily stores it; the instruction forwarding module parses the user input instruction and forwards it to the main analysis module; the main analysis module The real-time video is analyzed, and the analysis result is sent to the scheduling server module. The invention also provides a real-time video analysis method based on B / S architecture. The invention ensures that the frame fetching speed is not affected under high pressure, the communication efficiency between modules is higher than that of the traditional method, the communication error rate is reduced, and the playback progress of the server end is still aligned with the front end under high pressure.

Description

technical field [0001] The invention relates to the technical field of real-time video analysis, in particular, to a real-time video analysis system and method based on B / S architecture, and more specifically, to a B / S architecture-based system including cloud video speed and high pressure conditions A scalable real-time video analysis system and method for precise control and Socket cross-language communication under the invention. Background technique [0002] A real-time video analysis system based on B / S architecture refers to a system in which users can upload video to the server on the browser side, send analysis instructions, and obtain video frame analysis results in real time. For example, in the field of monitoring crowd analysis, users can upload monitoring video, play the monitoring video, and obtain analysis results such as crowd density, crowd movement grouping, and crowd attention in the current frame of the monitoring video. This kind of system occupies a la...

AI Technical Summary

Problems solved by technology

However, in the case of multiple users or multiple analysis items being used at the same time, if the packet header and segmentation identifier are broken into other packets, the data packet cannot be stored in the corresponding secondary buffer because the source of the data cannot be distinguished. Can be discarded, causing data corruption

However, in order to reduce the data damage rate, we can only increase the header density or identifier density, which greatly wastes transmission resources and reduces transmission efficiency.

Images