Image sampling algorithm used for external splicer or hybrid matrix

Abstract

The invention discloses an image sampling algorithm used for an external splicer or a hybrid matrix. The image sampling algorithm includes two parts, i.e., dowmsampling and upsampling, dowmsampling is to process an original 4:4:4 sampled signal to a 4:2:2 sampled signal, i.e., originally, a luminance component, a red chrominance component and a blue chrominance component are taken from each sampling point, and after processing, four luminance components, two red chrominance components and two blue chrominance components are taken from each four sampling points. Upsampling is a reverse process, i.e., to convert a 4:2:2 sampled signal to a 4:4:4 sampled signal. The image sampling algorithm used for the external splicer or the hybrid matrix is a simple algorithm, is relatively good in processing effect of an image, small in loss of details, and high in color quality; after downsampling, a requirement for channel bandwidth is lowered by one third, and the storage space and the occupancy rate of a CPU are also reduced by one third at the same time; additional modification of hardware is not needed, cost is saved, and the performance cost ratio is high.

Description

technical field [0001] The invention discloses an image sampling algorithm used for an external splicer or a mixing matrix, and relates to the field of electronic equipment manufacturing. Background technique [0002] The role of the external splicer / mixing matrix is ​​to realize the transmission and storage of images and videos. In the transmission and storage, due to the influence of the transmission channel bandwidth, memory size and CPU processing power, it is often necessary to up and down sample the image before processing. [0003] Traditional downsampling is to convert the original 4:4:4 sampling method to 4:2:2 sampling method. After downsampling, each pixel is converted from 24bit to 16bit, which can save transmission channel bandwidth, memory size and CPU processing power, and save one-third of resources. Upsampling is to convert each pixel from 16bit to 24bit for subsequent processing, which is the inverse process of downsampling. [0004] At each sampling poin...

AI Technical Summary

Problems solved by technology

[0005] However, after traditional downsampling and upsampling, since the color difference signal is simply decimated and interpolated, if the picture is too smooth, such decimation and interpolation will have little effect on the color of the image, but if you encounter an image with large contrast, Such extraction and interpolation will inevitably cause the degradation of image color quality and the loss of details.

After traditional downsampling processing, the image will inevitably lose details

The existing technology lacks an algorithm for image upsampling and downsampling, which can effectively minimize the image loss after downsampling, so as to meet the requirements of restoring the details of the image well after upsampling.

Images