Method of fast transcoding of surveillance video

Abstract

The invention discloses a method of fast transcoding of a surveillance video. The method is characterized by including the following steps: S100, determining that a to-be-transcoded source video fileas an input file, and determining a transcoded file as an output file; S200, detecting the type of the input file, wherein detecting includes the following specific steps: S201, reading and loading aconfiguration file, S202, judging whether the format of key-value pairs in the configuration file of the input file matches the format of key-value pairs in the step S201, if yes, executing a step S203, and otherwise, executing a step S204, S203, reading a second field of the input file as an address, reading contents in the address, judging whether the contents are the same as a first field of the key-value pairs in the step S201, if yes, executing a step S205, and otherwise, executing a step S204, S204, discarding the current input file, loading a next input file, and executing the step S202, and S205, using the first field, a second field and a third field of the current key-value pairs as parameters of the transcoder; S300, opening the input file; and S400, carrying out transcoding, and outputting the file.

Description

technical field [0001] The invention belongs to the field of monitoring video and relates to a method for fast transcoding of monitoring video. Background technique [0002] At present, there are many monitoring manufacturers on the market. Because there is no industry standard, each of them does their own thing, which leads to the complexity of the monitoring video format today, which makes it difficult for the subsequent playback preview and video analysis and processing of the video. The main performance is that it does not support universal players and does not support demultiplexing separation. , does not support video positioning, etc., and even some monitoring manufacturers cannot provide SDKs. The SDK functions of large manufacturers are relatively complete, while the SDKs of small manufacturers are different. It is extremely difficult to integrate and support all video formats. [0003] How to use a method to reduce the difficulty of playing and analyzing in the bac...

AI Technical Summary

Problems solved by technology

As we all know, video encoding is time-consuming and resource-intensive. If it takes too much time to transcode before video playback or analysis, it is not worth the loss. At the same time, transcoding should support as many different video input formats as possible. Some transcoding methods have the following limitations:

[0004] 1. Poor compatibility. By directly integrating the manufacturer's SDK functions, the more the SDK is collected and integrated, the stronger the transcoding compatibility will be. Manufacturers that do not provide the SDK can hardly

[0005] 2. The processing speed is slow. Through the decoding function provided by the manufacturer's SDK, and then re-encoding to obtain a standard video format, but encoding is time-consuming and resource-intensive

[0006] 3. The device is highly dependent and costly. In order to improve the efficiency of encoding work, GPU can be used for acceleration. The price of GPU devices is not high, the speed increase is not obvious at the low end, and the high end is too expensive, and there are few supported video formats.

[0007] When encoding video, it is not easy to balance image quality, space occupation, and time-consuming encoding and decoding. Generally speaking, if you want good image quality, you either need to occupy a large storage space, or the compression rate is high but the algorithm is complex and time-consuming, and decoding is also difficult. so;

Images