Bandwidth optimization method and device

Abstract

The invention provides a bandwidth optimization method and device. The bandwidth optimization method comprises the steps that parameter range information of an image is determined according to parameter information of the image and hardware parameter information of an image processing device, wherein the parameter range information comprises parallax error range, width of the image, memory digits and / or the weight of an abnormal parallax value; the weight of a normal parallax value is calculated out according to the parameter range information and a preset calculation rule; basic data structure information of the image is built according to the weight of the normal parallax value and the weight of the abnormal parallax value, wherein the basic data structure information is in an integral type. By means of the technical scheme, the bandwidth can be optimized.

Description

technical field [0001] The present invention relates to the technical field of image display, in particular to a bandwidth optimization method and a bandwidth optimization device. Background technique [0002] At present, most of the methods that can calculate disparity more accurately include aggregation and scanline optimization modules, but in modules such as aggregation and scanline optimization, because the algorithm involves a large number of memory reads and writes, it requires high bandwidth. The high bandwidth hinders the real-time calculation of disparity. The inability to calculate parallax in real time makes it impossible to popularize and use a large number of applications, such as viewing 3D videos with naked eyes and real-time 3D reconstruction. [0003] In order to reduce the bandwidth of the disparity calculation method, people have done a lot of exploration. For example, without aggregation processing or scanline optimization, this method increases the er...

AI Technical Summary

Problems solved by technology

For example, without aggregation processing or scanline optimization, this method increases the error rate of disparity calculation, and it is found that there are obvious disparity calculation errors, and the algorithm fails; by adding additional hardware to realize disparity calculation, the cost and equipment will be reduced after adding hardware. The heat dissipation has risen sharply, which cannot be popularized; reducing the parallax range requires only a small angle between the cameras during the shooting process, which can only be adapted to the parallax calculation of individual cases, and the application range is small, so it cannot be popularized; The previous method of reducing bandwidth through data compression requires additional space, and the logic is complex, and the bandwidth optimization space is small, and the algorithm still cannot be popularized.

Images