Applications of spatial varying filter on image interpolation, demosaicing and video denosing [i.e. denoising
by Chan Tai Wai
M.Phil. Electronic and Computer Engineering
x, 96 leaves : ill. (some col.) ; 30 cm
With the advance of computer networking technologies, multimedia applications are becoming more and more popular. Among many different kind of multimedia, image and video are undoubtedly two of the most important ones. Consequently, image and video processing techniques are getting more and more important....[ Read more ]
With the advance of computer networking technologies, multimedia applications are becoming more and more popular. Among many different kind of multimedia, image and video are undoubtedly two of the most important ones. Consequently, image and video processing techniques are getting more and more important.
Many algorithms have been developed to tackle image and video related problems. Spatial varying filter (SVF) is one of the tools that developed to handle some image and video problems. One of the main advantages of SVF is its simplicity, which implies ease of implementation. Thus, it is suitable to apply SVF in many handheld devices like mobile phone. SVF could be applied in many areas of multimedia processing. Three SVF applications would be discussed including image interpolation, demosaicing and video denoising.
Image interpolation or image zooming is a very common image processing technique, and its usage is to interpolated an image with a small spatial resolution to an image with a larger spatial resolution. By using SVF, missing pixel values in the interpolated image would be constructed by taking a suitable weighted average of their neighboring pixel values.
Demosaicing or color interpolation is a well-known problem in many low end digital cameras (DC) which using only one charge-coupled device (CCD) module combined with a Bayer color filter. The aim of demosaicing is to generate the missing color pixel values and to construct a full-color resolution image. By applying SVF, each missing color pixel value would be generated by exploiting the characteristic of its neighboring color pixel values.
Video denoising is a process to remove unwanted information or noise from a noisy video clip. This could improve not only the video perceptual quality but also the compression efficiency of some video encoders. Good results could be obtained by employing SVF for video denoising.