Buffer Apparatus for Video Encoding Memory Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional buffer devices face challenges in suppressing the size of the buffer memory required to store reference image data temporarily and the data transfer rate of external memory, particularly when encoding high spatial frequency images like 4K or 8K, leading to increased scale, size, and power consumption, as well as deterioration in motion search precision and image quality due to insufficient data transfer time.
Innovation Solution
A buffer device with a memory and buffer control unit that stores reference image data for each search range and controls the retention and updating of data based on the position of encoding target blocks, allowing reference image data from previous search ranges to be retained and reused, and generating data for protruding pixels, thereby optimizing memory usage and transfer rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the reference image buffer stores reference image data for the entire search range (image width × search range height), then the data transfer rate required of the external memory is reduced, but the size of the reference image buffer increases massively
Solution Approach 1:
The patent divides the reference image buffer into multiple buffer units corresponding to different buffer regions (first buffer region, second buffer region, third buffer region). Each buffer unit stores reference image data for a specific region independently. This segmentation allows the system to transfer only the necessary buffer regions from external memory rather than the entire buffer, reducing the data transfer rate requirement while maintaining adequate buffer size for each region.
2Volume of stationary object
If the reference image buffer stores only reference image data in the search range of the encoding target block, then the buffer size is reduced, but the data transfer rate of the external memory increases
Solution Approach 1:
The patent pre-loads reference image data into the reference image buffer before encoding operations begin. By having the buffer already populated with necessary reference data from external memory in advance, the system avoids the need for high-speed data transfer during the actual encoding process. This preliminary action allows the buffer to be relatively small while still meeting data transfer requirements during encoding.
3Productivity
If the reference image buffer stores reference image data for high spatial frequency images (4K or 8K), then the coding efficiency is maintained, but the scale, size, and power of the LSI increase
Solution Approach 1:
The patent implements different buffer size configurations for different regions of the reference image. The first buffer region has a different size than the second and third buffer regions, optimized according to the specific requirements of each region. This local quality approach allows the system to maintain high coding efficiency for high spatial frequency images by ensuring adequate buffer capacity where needed, while reducing overall buffer size and LSI complexity by not uniformly increasing buffer capacity across the entire image.
Data Source
AI summary
A buffer device stores, in relation to a first area and a second area, which are partial areas respectively located on an identical line of an encoding target frame, and a third area, which is a partial area on the line below the identical line, reference image data for each search range of a reference frame that is referred to during motion search processing executed in order of the first area, the second area, and the third area, the buffer device including a memory for storing the reference image data, and a buffer control unit which, when the motion search processing is executed on the second area, executes control so that among the reference image data included in a first search range that is referred to during the motion search processing of the first area, the reference image data included in a third search range that is referred to during the motion search processing of the third area are left in the memory without being deleted.