Research And Application Of Phase Correlation Algorithm Based On Kalman Filter In Image Mosaic

Image mosaicing is a technique that involves combining multiple images with overlapping areas to create a large,high-resolution image.This technology has extensive applications in various fields,such as remote sensing image processing,medical imaging,and more.This article explores the significance and potential of image stitching technology in the application of scanning pens.Scanning pens are portable tools for collecting text and images,widely used in document scanning,archives management,and other related fields.However,due to the low image resolution of scanning pens,traditional image stitching methods cannot properly register during registration,resulting in the inability to stitch images together,while also consuming a significant amount of processing time.This article uses phase correlation algorithm to perform image registration,which,because it uses Fourier transform to process images in the frequency domain,can effectively handle low-resolution images and greatly accelerate processing speed.In addressing the noise problem in registration offset,this article introduces the Kalman filter algorithm,which significantly reduces character ghosting in registration results and effectively improves the quality of stitched images.The main work of this paper is as follows:(1)In response to the shortcomings of traditional phase correlation algorithms,improvements are made.Traditional phase correlation algorithms cannot judge the reliability of registration results during image registration.In this paper,the traditional phase correlation algorithm is further improved by judging the reliability of the calculation results and discarding the registration failure results.This algorithm can effectively reduce the problem of missing strings during stitching and improve the effect of image stitching.(2)In response to the noise problem in registration results,the Kalman filter is introduced.In order to reduce the noise in the registration results and make the stitched image smoother,this paper introduces the Kalman filter.Based on the hand motion model,a linear motion model is established to filter the registration offset.Experimental results demonstrate that this method effectively reduces the problem of ghosting in stitched images and improves the quality of image mosaicing.Experiments show that scanning pen image stitching algorithm is proposed in this paper,on the basis of guarantee the speed of image matching,improve the quality of the image splicing.