Cfa-based Image Reconstruction Algorithm

Digital imaging technology is a set of optical, electrical, computer anddigital image processing technology and other disciplines in an integratedhigh-tech, but also Digital imaging technology is all important research field incomputer vision. Related techniques have been broadly applied in videomonitoring, human computer interaction, intelligent transportation system,military field and so on. Digital imaging technology belongs to the low butfundamental operation of computer vision. An effective algorithm plays a vitalrole to complete the follow-up work. Therefore, the research and development ofdigital cameras has a very real sense.The arrangement of the color filter and the image reconstruction algorithmare the two key points in the digital imaging technology and also the hot issue.This thesis mainly focuses on digital imaging technology based on the CFA(Color Filter Array). A certain amount of work has been done mainly in the colorfilter array arrangement and image reconstruction algorithm.Firstly, the arrangement of color filter has a great impact on both theobjective and subjective quality of image reconstruction, full-color image. Threedifferent color filter arrays are presented based on the spectral structure of colorfilters. Experiments show that this CFA can effectively reconstruct the colorimage.The classical image reconstruction algorithm are discussed and conductedby the thorough research. A compare is made about the advantage anddisadvantage of the various algorithms from the subjective and objective.Due to the artifact of classical image reconstruction algorithms in textureregions, this paper proposes an adaptive region demosaicking usingMulti-Gradients. It first recovers green components according to the operatorsand the texture regions. Then the chrominance channels are reconstructed basedon the interpolated green components and constant-hue rule. Experiments provethat this algorithm can improve the effectiveness from the subjective andobjective.