Bilateral filter

About: Bilateral filter is a research topic. Over the lifetime, 3500 publications have been published within this topic receiving 75582 citations.

Image noise reduction apparatus and method, recorded medium recorded the program performing it

Abstract: An apparatus and a method of reducing a noise component existed in a surrounding part of an image photographed in an image sensor. According to an aspect of the present invention, an apparatus differentially reducing a noise of an image according to an area of the image includes a filter area selecting unit, selecting a filter area including a plurality of adjacent pixels based on an object pixel; and a noise reducing unit, computing pixel data of the object by using pixel data of the plurality of adjacent pixels in the filter area, whereas the filter area selecting unit determines the size of the filter area according to the distance between the object pixel and a center pixel of the image. The present invention can prevent a noise component of a surrounding part from being amplified by using a different low pass filter in a center part and a surrounding part of an image and can acquire the resolution and quality of the desired image.

Classified quadtree-based adaptive loop filter

Abstract: In this paper, we propose a classified quadtree-based adaptive loop filter (CQALF) in video coding. Pixels in a picture are classified into two categories by considering the impact of the deblocking filter, the pixels that are modified and the pixels that are not modified by the deblocking filter. A wiener filter is carefully designed for each category and the filter coefficients are transmitted to decoder. For the pixels that are modified by the deblocking filter, the filter is estimated at encoder by minimizing the mean square error between the original input frame and a combined frame which is a weighted average of the reconstructed frames before and after the deblocking filter. For pixels that the deblocking filter does not modify, the filter is estimated by minimizing the mean square error between the original frame and the reconstructed frame. The proposed algorithm is implemented on top of KTA software and compatible with the quadtree-based adaptive loop filter. Compared with kta2.6r1 anchor, the proposed CQALF achieves 10.05%, 7.55%, and 6.19% BD bitrate reduction in average for intra only, IPPP, and HB coding structures respectively.

Efficient perceptual attentive super-resolution

Abstract: An efficient Perceptually Attentive (PA) Super-Resolution method is proposed to significantly reduce the computational complexity of iterative super-resolution algorithms without loss of the desired perceptual quality. A perceptually significant constrained set of active pixels is selected for processing by the SR algorithm based on a just noticeable distortion threshold model. These selected active pixels are further reduced by using saliency information that is determined by a visual attention model. Furthermore, the active pixels lying in the attended regions are processed at a higher accuracy by the SR method relative to pixels in other regions. Simulation results are presented to show the preserved desired visual quality and a 30–40% reduction in complexity over a highly efficient Selective Perceptual Fast Two-Step (SELP-FTS) scheme.

Sharpening filter for images with automatic adaption to image type

Abstract: A method of filtering images with a filter that is dependent on image type. For binary images, a fixed Laplacian filter is used for all pixels. For non-binary images, an automatic level-sensitive Laplacian filter is used. For non-binary images, Laplacian filter coefficients and normalization divisors are made a function of intensity on a pixel by pixel basis. Pixels indicating a high intensity may be greatly changed by neighboring pixels. Pixels indicating a low intensity receive relatively little or no filtering. In a specific example embodiment, the intensity range is divided to bands, with a different filter for each band. The resulting images are visually sharpened (edge enhanced) with very little impact in darker areas of the image.