Research On The Color Interpolation Algorithm Of CMOS Image

CMOS image sensor with monolithically integrated, high integration, random access,read speed, low power consumption, strong anti-radiation, cheap cost and other advantages,making its wide range of applications in the consumer electronics, medical, industrialproduction monitoring, aerospace, military and other fields. The color interpolation of CMOSimage is an important part of the digital imaging system. a good color interpolation algorithmcan improve the clarity, color reproduction and sharpness of the image, resulting in bettervisual effect.In the study of several commonly used classic color image interpolation algorithms,including nearest neighbor interpolation algorithm, bilinear interpolation algorithm, based onthe calculation of the first-order differential and second-order differential edge interpolationalgorithm and Adams-Hamilton adaptive interpolation algorithm.Taking into account the influence of noise in the CMOS image sensor data acquisitionand transmission process, So this paper proposes an improved interpolation algorithm whichcan filter out some of the noise. Calculating double gradients value to judge the location ofthe noise point in the restoration of the green component and to further remove noise.According to both the channel correlation and the pixel correlation, the principle is based oncolor difference theory in the restoration of the blue and red component.Simulating the common nine non-iterative interpolation algorithm applied to thehardware system, the experiments show the algorithm has advantages in simple principle,good interpolation effect, fast speed, better performance than traditional algorithm to someextent and so on.