Adaptive Color Filter Array Interpolation Algorithm Based on Hue Transition and Edge Direction

Most of the digital cameras used now has single sensor with an array of color filters to capture the digital images. Color Filter Array (CFA) uses alternating color filters to sample only one color of a pixel location to reduce the price. In the entire image, only one third of sampled pixels have genuine values from the sensors. Remaining pixels are interpolated from these genuine values. In this work, a new adaptive CFA interpolation model is proposed. Pixels are divided into edge pixels and non-edge pixels. If the pixel is not in an edge, then smooth hue transition interpolation is used. For the pixels in the edge, Edge sensing interpolation algorithm is used. A mathematical model is proposed for interpolation based on the direction of the edges. Two sample images are taken and reconstructed by methods such as hue transition interpolation, edge sensing interpolation and combined approach. PSNR is used to compare the results of proposed technique.