Data lockless caching method for video monitoring equipment and server

Abstract

The invention provides a data stream lockless caching method for video monitoring equipment and a server, and the method comprises the steps: initializing a cache region which comprises an index region and a data region; when a data stream is collected, packaging the data stream to obtain packaged data; when the length of the frame data in the packaged data exceeds the preset maximum single-framelength, discarding the packaged data; when the length of the frame data in the packaged data does not exceed the preset maximum single-frame length, storing the packaged data into the data area, and storing the start-stop position information of the packaged data in the data area into the index area; and when an instruction for acquiring data from the cache region is received, acquiring corresponding packaged data in the data region through the start-stop position information stored in the index region, and sending the packaged data. According to the method, system resource calling is greatlysaved, and the data caching efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of video monitoring, in particular to a data lock-free caching method and server for video monitoring equipment. Background technique [0002] Existing video surveillance equipment caches data using chained splicing caches. When data streams are generated, the required memory is temporarily requested from the system dynamically and connected to the end of the previous storage node to form chained storage. Due to the "consumption" cycle Uncertain and irregular existence of multiple "consumers", while "producer" data is continuous, it is easy to cause data accumulation, resulting in a large amount of memory not being released in time. [0003] Among them, the producer of data refers to the electronic device that generates data, for example, the producer of the camera when the camera collects video data stream; the consumer of data refers to the electronic device that uses the data, for example, displaying surv...

AI Technical Summary

Problems solved by technology

[0004] In addition, in order to ensure the correctness of the read and write data, a mutex is added between the read and write threads for isolation operations. On average, a lock / unlock operation is performed every 30ms, which reduces the utilization efficiency of the system CPU. At the same time, it also repeatedly Dynamically apply for / release memory, resulting in a large number of memory fragments, resulting in low memory usage efficiency, which in turn affects the instability of product storage modules and preview modules

Another solution uses a fixed-size circular buffer queue for recycling, but since the size of each frame of data is not fixed during the actual operation of the product, this method will lead to waste of memory

These have greatly reduced the efficiency of data cache usage, leading to the problem of low efficiency of data cache

Images