Method for controlling video coding flow rate of monitoring video sending end

Abstract

The invention discloses a method for controlling the video coding flow rate of a monitoring video sending end. The method mainly comprises the steps of (1) collecting a real-time available bandwidth data set of a video sending scene network; (2) constructing a simulation training environment of a monitoring video sending end by utilizing the real bandwidth data, determining the highest available bandwidth for sending the monitoring video in real time by the training environment according to the real bandwidth data, taking the highest available bandwidth as the sending rate of the video, and receiving the code rate selected by the deep reinforcement learning model to adjust the encoding code rate of an encoder; (3) constructing a continuous action output deep reinforcement learning model based on a trust region, and training the model by utilizing a simulation environment; (4) putting the trained model into a monitoring video, integrating the monitoring video into a real environment forinteraction, and carrying out online training optimization; and (5) integrating the optimized deep learning model to a monitoring video sending end, and making an encoding rate decision of an encoderof the sending end. According to the invention, the coding flow rate control problem of the monitoring video sending end is solved by using deep reinforcement learning.

Description

technical field [0001] The invention relates to the field of real-time video transmission, in particular to a method for controlling the video coding flow rate of a monitoring video monitoring terminal. Background technique [0002] Surveillance video usually has high requirements in terms of real-time, fluency, and video picture quality. However, in the actual monitoring environment, from the monitoring video collection end to the receiving end taking the monitoring room as an example, the video transmission often passes through a relatively complex network environment, which will lead to bandwidth limitations and delay fluctuations. The situation occurs, thereby affecting the real-time performance, fluency and clarity of the monitoring video playback end (receiving end). In order to ensure the transmission effect of surveillance video and improve the viewing experience of surveillance video, it is necessary to optimize all links in the surveillance video transmission proc...

AI Technical Summary

Problems solved by technology

However, in the actual environment, since the change of the actual available bandwidth is usually irregular, it is quite difficult to estimate the actual available bandwidth at the next moment, so it can only be observed by observing some measurable characteristics during the video transmission process Parameters to make a rough estimate of the current transmission environment, and select the code rate of the encoder at the next moment according to the observed characteristic parameters

Images