Image sharpening method and device

Abstract

The invention provides an image sharpening method and device. The sharpening method comprises a converting step, a filtering step and a coring and denoising step, all of which are carried out in sequence. The method comprises the steps that neighborhood data are formed according to to-be-processed pixel points, a first pixel point on the left, a first pixel point on the right, a second pixel point on the left and a second pixel point on the right, after frequency domain converting is carried out according to the neighborhood data, calculation is carried out to obtain coefficient vectors, the sub-filtering weighting coefficient is calculated according to the coefficient vectors, data obtained after filtering are obtained according to processing of the neighborhood data, the sub-filtering weighting coefficient and a sub-filtering pulse response function, a sharpening weighting value is worked out according to the neighborhood data, the sub-filtering weighting coefficient and a weight adjusting parameter, denoising is carried out on the data obtained after filtering according to a threshold value parameter, sharpening is carried out on the denoised data, and the denoised data are processed to form sharpened data. Sharpening processing is carried out in a self-adaptation adjusting mode according to the image content, and accordingly the purpose that the sharpened visual effect is more natural is achieved.

Description

Technical field [0001] The invention relates to the field of digital video processing, and in particular to an image sharpening method and a sharpening device using the sharpening method. Background technique [0002] With the widespread use of computers and the emergence of digital TVs, people have higher and higher requirements for the definition of video images. When standard-definition video is enlarged to high-definition video playback, the details of the edge of the video image will be lost, making The overall image is blurred. The clarity and sharpness of the image will also be lost due to encoding and various processing such as noise reduction, de-ringing effect, and de-blocking. At this time, edges and details are usually enhanced to recover lost high-frequency information or increase high-frequency information to improve image resolution and perceived clarity. This post-processing technique is called sharpening. [0003] Existing sharpening processing solutions usually ...

AI Technical Summary

Problems solved by technology

[0004] However, the filter gain in LTI processing is fixed and cannot be adjusted according to the actual content of the image, and multiple sub-filters will be used in Peaking processing, and the gain of the sub-filters is also fixed and cannot be adjusted adaptively

Since the processing method is single, and different areas of the image require different processing methods, using a fixed filter makes the sharpening result unnatural

In addition, the filter frequency ranges in LTI and Peaking overlap, and the image data in the overlapping range will be sharpened multiple times, which may produce an edge gain superposition effect, thereby reducing the visual effect of the image

Images