Modified Adaptive Color Plane Based Color Interpolation Algorithm for Digital Image Sensors

At present, there is only one image sensor and a color filter array covering its surface to collect color images in most imaging devices. Each pixel of the collected color image can only obtain one color component from digital image sensor directly, and the other two color missing components need to be calculated through color interpolation. Since Bayer-type color filter array widely used in various kinds of sensors has good color signal sensitivity and color recovery characteristic, we reconstruct the two missing color components in Bayer mode images in this paper according to it. Two Color Interpolation algorithms widely used in digital imaging devices are introduced in detail in this paper, Gradient Edge-oriented (GEO) Interpolation Algorithm and Adaptive Color Plane (ACP) Interpolation Algorithm, and then we present a modified adaptive Color layer Interpolation Algorithm based on them. In the experiments, 10 images are selected from the IVC database. 40 sets of data are generated from the images and then they are used to analyze from Peak Signal to Noise Ratio (PSNR), Structural Similarity (SSIM), Feature Similarity (FSIM) and Computing Time quality metrics to study the quality of the 10 images after the processing of interpolation. This four metrics are the basis of judging the processing effect of MACP, GEO and ACP algorithms. The results of the experiments show that Modified adaptive Color plane (MACP) algorithm can solve the contradiction effectively between the reconstruction of high-quality color images and reduction of computational complexity, and improve the processing speed of color image reconstruction exactly.