Research Of Panchromatic CFA And Interpolation Algorithm Based On Fourier Transform

With the rapid development of digital technology, digital cameras gradually come into people’s daily life, and become more and more important in many fields. So the image quality of digital cameras is paid more and more attention to. In order to reduce cost and size, only one image sensor overlaid by a color filter array is used in most digital cameras, such that only one color is sampled at each pixel of the image sensor. To obtain the full color image, the information of neighbor pixels is used to estimate the missing color information, this process is commonly referred to as color interpolation or demosaicking. The quality of the full color images is closely related to the color filter array and the corresponding color interpolation algrothim. In order to find some new methods to improve the quality of the color images, the color filter array and color interpolation algorithms are studied in depth in this paper.First, some commonly used color interpolation algorithms based on Bayer CFA are analyzed. According to the characteristics that the CFA sampling image could be interpreted as the luminance component lacated in baseband and the color difference components located in high band, a new color interpolation algorithm based on Bayer CFA is proposed to overcome the shortcomings of the existing algorithms in this paper. In the new method, the Fourier analysis technology is used to design the modulation signals of red-green color difference and blue-green color difference. The CFA sampling image signals modulated by the above modulation signals contain only one color difference component in baseband, and then the modulated signals are filtered with a low-pass filter to obtain the information of the red-green color difference and the blue-green color difference. Finally, the information of red, green and blue color component is recovered using the linear method, and thus the full color image is recovered. Compared with the commonly used color interpolation algorithms based on Bayer CFA, the proposed method outperforms all of them in the sense of the interpolation effect, especially at the edges and the detail areas of the image.Moreover, there is aliasing between the luminance component and the color difference components in the spectra of the images sampled with various commonly used CFA patterns, and it reduces the quality of the reconstructed full color image. In order to improve the image quality, a new method based on Fourier analysis technology is proposed to design a new kind of CFA pattern, the corresponding color interpolation algorithm is also studied. The principle of the method is to separate the color difference components and the luminance component as far as possible in Fourier domain. Meanwhile, the aliasing between the color difference components modulated to different frequency should be also reduced. Some new CFA patterns are designed using the new method, and the aliasing is reduced in the spectra of the images sampled with the new CFA patterns. In the Matlab simulation, the advantages of the new CFA are proved by means of subjective visual perception and objective evaluation index (i.e. CPSNR and VIF). The quality of the full color images reconstructed with the new CFA patterns and the corresponding color interpolation algorithm is much higher than the previous ones.