Image dithering processing method and device

Abstract

The invention discloses an image dithering processing method. The image dithering processing method comprises the following steps that: image data are acquired; based on a No. N pixel value in the image data, a No. N random array is generated, wherein the N is an integer greater than 0; according to the No. N random array, a No. N + 1 pixel value is processed by dithering, and when the No. N + 1 pixel value is processed by dithering, a median of the No. N + 1 pixel value is generated; according to the median of the No. N + 1 pixel value, a No. N + 1 random array is generated; and according to the No. N + 1 random array, a No. N + 2 pixel value is processed by dithering, until pixel values of the image data are processed. The method can keep the correlation between adjacent pixels, and can effectively eliminate the color bar problem and other problems while maintaining consistency of adjacent pixel colors. The invention also discloses an image dithering processing device.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image shaking processing device and an image shaking processing method. Background technique [0002] Dither (jitter) is an intentional noise used to randomize the quantization error to prevent problems like Banding (color band) caused by greatly stretching the image. At the same time, through the dithering process, the detail components of the image can be increased, the image can be sanded, and the water ripple of the image can be eliminated. [0003] Due to the limitation of the pixel width of each point in the image, a large number of algorithms for compensating the truncation error have been generated, that is, random numbers are generated for comparison with the pixels of each point in the image, and corresponding pixel values ​​are generated through different algorithms, that is, after the dithering is generated. new image. [0004] At present, the mo...

AI Technical Summary

Problems solved by technology

[0006] The disadvantage of the existing technology is that the approximation of the value based on the comparison with the selected matrix only increases the details of the pixel without considering the influence of adjacent pixels, or only generates a fixed value based on the comparison of the matrix generated by experience, thus generating The difference between the new color and the adjacent image may be very large, which is prone to color noise

Images