Method of reducing block and mosquito noise (effect) in images

Abstract

The present invention provides an image clutter removal method that can appropriately remove both block clutter and mosquito clutter without affecting picture quality. For the brightness component image data, first brightness component image data that smoothes the boundaries between blocks is created, and then second brightness component image data that is overall smoothed is created based on the first brightness component image data, and then Edge image data is created by subtracting the second brightness component image data from the first brightness component image data, and then corrected edge image data is created that corrects the edge image data based on predetermined conditions, and finally the second brightness component image data is created. The third brightness component image data is obtained by adding the component image data and edge image data. Regarding the color difference component image data, first color difference component image data that is overall smoothed is created from the color difference component image data. Then, the third brightness component image data and the first color difference component image data are output as the final output.

Description

technical field [0001] The present invention relates to an image clutter removal method for removing block clutter and mosquito clutter generated when decoding an image compressed and coded in block units. Background technique [0002] As a high-efficiency image data compression method, a method of realizing high compression by orthogonal transform coding is generally used. If you want to make a brief description of this method, that is to say: first, use the block circuit to divide the input image signal into blocks of block size such as 8×8 pixels, and use an orthogonal exchange circuit to Orthogonal switching is performed on the divided blocks to generate frequency component signals. Next, linearly quantize the orthogonally exchanged data through the quantization circuit with the determined quantization step size. Then, a variable length code is assigned to the quantized result by a variable length coding circuit to generate a coded image signal. As the orthogonal tran...

AI Technical Summary

Problems solved by technology

[0008] However, in the method for removing block clutter described in Patent Document 1, sometimes it cannot be determined whether it is a boundary between blocks where a large difference in pixel value between adjacent pixels occurs due to generation of block clutter. Or is the large difference in pixel value between adjacent pixels just the outline in the image? If the outline in the image is mistaken for the boundary between blocks, and after block clutter removal processing, the outline in the image will be blurred, which may cause the image quality to deteriorate

[0009] In addition, the mosquito clutter removal method described in Patent Document 2 is a simple filtering process, so it is certain that no processing time is required. However, due to its simplicity, it is impossible to properly remove mosquito clutter without impairing the image quality. wave, still blurring the outline in the image, resulting in the possibility of image quality degradation

[0010] In addition, as long as quantization errors caused by quantization and inverse quantization still exist, block clutter and mosquito clutter should appear in corresponding parts of the reproduced image, even if the degree is different. In spite of this, there is currently no technology that satisfies both block clutter and mosquito clutter removal

Images