Research Of Image Quality Optimization Technology For Digital Still Camera

With the development of science and technology, the application ofdigital still camera is become more and more wider. From the record oflittle drops of life to the beautiful photos at hand taken by the digital stillcamera, and from the video calls between friends to the site monitoring ofpublic places, the digital still camera has integrated into every corner ofthe family and the society. The demanding of the camera image quality bypeople has become increasingly higher. The majority of the currentlyavailable digital still camera uses the cmos image sensor, and the processfrom the image data exported by the senor to a complete RGB image dataneeds a series of processing. This paper analyzes some factors whichinfluences the image quality of the digital camera, and studies thecorresponding solution methods, improving the color interpolationalgorithm and the auto white balance algorithm, and the above algorithmhas been applied to the digital camera exploited by the company ofdushen.According to the Bayer format data generated by the image senor inthe digital camera, we improved the existing color plane interpolationalgorithm. The algorithm firstly improved the classifier of the brightnesschannel and the chroma channel, refined the edge classification, and thenoptimized the existing formula of interpolation estimation, so that theestimated value is more precise. The improved algorithm has an accurateresult for the reduction effect of the mutant edge, and has a lowcomputational complexity degree at the same time.For the problem of the inaccurate automatic white balance of digitalstill camera and the not enough real of the color reduction, an automaticwhite balance algorithm is proposed based on the LAB color space.Because the color perception is linear in the LAB color space, that is tosay, the same number of variations on the color space should produceapproximately the same visual importance of changes, therefore, it has agood effect to have color reduction in the LAB color space. Thealgorithm according to the color space conversion, first makes the sRGBgraphics convert to the LAB color space. Then, the color correction is made for the A,B color components using the gray world model in LABcolor space. Experimental results show that it can get a good effect ofcolor reduction in the LAB color space using the automatic whitebalance.