Research On Underwater Image Mosaic Algorithm Based On SIFT

At present,the competition for marine resources is increasingly fierce in the world,and underwater image processing technology has an important application in the exploration and development of seabed resources,so the requirements for underwater image processing technology are further improved.On the basis of studying underwater image mosaics,this paper studies underwater image enhancement processing,feature point positioning and matching,image mosaics algorithm and other aspects.The main work is as follows:(1)Aiming at the problems of low definition and low contrast in underwater images.In terms of sharpness improvement,average,gaussian and bilateral filtering algorithms are adopted to realize image denoising.In terms of contrast improvement,histogram equalization is adopted to stretch the gray level of the original image to improve the contrast between different objects in the image.Three image quality evaluation indexes,mean square error,peak signal-to-noise ratio and one-dimensional entropy,were used to verify that no image distortion occurred in the original image after enhanced processing.(2)Aiming at the problems that the registration accuracy is not high and the registration time is long in the underwater image registration process.RANSAC algorithm was used to realize rough matching of feature points based on SIFT operator,and the homography matrix was iteratively updated by using the nearest neighbor principle and interior point criteria,so as to improve registration accuracy and reduce registration time.(3)Aiming at the problems that are easy to occur in the process of underwater image fusion by weighted average method,such as the reduction of definition and the jointing.The resolution of the enhanced image was stratified based on the Laplace pyramid method,and the fusion images of different layers were successively fused by the weighted average method,and the fusion images of different layers were integrated into a mosaic image,so as to realize the seamless fusion of the underwater image with high resolution and avoid the double shadow phenomenon.This work systematically studies underwater image mosaic based on image enhancement and improved RANSAC algorithm.By analyzing the enhancement,registration and fusion effects of different kinds of experimental images,the underwater image enhancement,feature point positioning and matching,and high resolution seamless stitching were completed,which improved the accuracy and stability of the algorithm.At the same time,this work can also provide important value for the future development of underwater image processing technology and the use of image technology to detect seabed resources.