Method and device for allocating encoding channel memory

Abstract

The present invention relates to a method and device for allocating encoding channel memory, comprising: dividing a public buffer pool into multiple buffer areas; determining one of the multiple buffer areas in the public buffer pool as the first channel The memory unit of the first reconstructed frame; when it is determined that the original reference frame corresponding to the first channel is used as the first reference frame, the memory unit of the first reconstructed frame is released to the public buffer pool; when it is determined that the first reconstructed frame is used as the first reference frame When the frame is used as the first reference frame, the memory unit of the original reference frame corresponding to the first channel is released to the public buffer pool; the memory unit released from the previous channel to the public buffer pool is determined as the second reconstruction corresponding to the current channel The memory unit of the frame; when it is determined that the original reference frame corresponding to the current channel is used as the second reference frame, the memory unit of the second reconstructed frame is released to the public buffer pool; when it is determined that the second reconstructed frame is used as the second reference frame , release the memory unit of the original reference frame corresponding to the current channel to the public buffer pool.

Description

technical field [0001] The present invention relates to the technical field of video and image processing, in particular to a method and device for allocating encoding channel memory. Background technique [0002] Compared with traditional analog video recorders, digital video recorders (Digital Video Recorders, referred to as "DVR") use hard disk recording, so they are often called hard disk recorders. It is a computer system for image storage and processing, with functions of long-term video recording, audio recording, remote monitoring and control of image / voice. And its codec system is generally self-compiling and self-decoding, and the number of channels and resolutions of encoding and decoding are relatively fixed. The more codecs, the more memory it takes up, and the higher the memory cost. [0003] The memory occupied by each channel of the current encoding system is independent, mainly because the reference frame and the reconstructed frame occupy more memory. Dur...

AI Technical Summary

Problems solved by technology

From the user's point of view, the encoding of the 16 channels works in parallel, but from the perspective of the encoding hardware unit, the encoding of the 16 channels actually works serially, that is, the hardware only writes the encoding of one channel at a time. Reconstruct the frame. At this time, the memory of the 45M reconstructed frame of the other 15 channels is in an idle state, resulting in a large memory usage and increasing the memory cost.

Images