Research On Bayer Image Color Reconstruction Algorithm Based On Gradient Weighted Residual Interpolation

Due to constraints of size and cost,camera manufacturers mostly choose a single CCD or CMOS image sensor structure,and then cover the sensor surface with a layer of color filter array.So that the final imag of such camera is a Bayer image which composes three color blocks(red,green and blue).In order to ensure the output of a full-color image,the corresponding interpolation algorithm must be used,which can add the other two color components to each Bayer image pixel.This process is called color reconstruction.Existing literatures have proposed some reconstruction algorithms,but they have disadvantages such as poor edge processing effect,zipper effect,block effect,and false color.In view of these problems,this paper proposes a gradient weighted residual interpolation algorithm.This algorithm first uses linear interpolation to obtain the initial estimates of all pixels from both horizontal and vertical directions,and then obtains the color difference estimates of all pixels in horizontal and vertical directions.The Bayer image plane,the linearly interpolated Bayer image plane,and the residual plane are used to jointly provide weights in four directions,and then the weight values are combined with the estimated color difference values,and then added to the original red and blue channel values to restore the complete green plane.Finally,the reconstructed green component is further refined based on the color difference plane.Different ambient light will cause color shift in the image obtained after interpolation.Color correction is needed for this problem.Aiming at the problems of missing image details and low color saturation in the existing color correction algorithms,an image color correction algorithm based on improved gain coefficients is proposed.This algorithm first corrects the color cast components in the image by using the gray world method with improved gain coefficients,and then converts the processed image from RGB color space to HSI color space.The saturation component is linearly stretched,and then the image is converted back to the RGB color space.Finally,the image is adjusted to obtain the final color corrected image.The experimental results show that the proposed algorithm effectively solves the problem of loss of texture details and edge blurring of the image.Compared with other algorithms,this algorithm can better solve the problem of color cast in the image,and the brightness and color saturation of the processed image are closer to the real scene.