Automatic image registration algorithm based on improved Harris corner detection

Image registration, as an important part the computer image processing, has been widely used in such technologies as video compression, 3D reconstruction, image restoration, etc. The paper proposes a new improved Harris algorithm based on corner detection. Based on the characteristics of changes in gray gradient of corners, the paper puts forward the concept of level of similarity of image scale pixel. By detecting the level of similarity between the gray value of the center point of the window and other pixels in the neighborhood, the paper accurately wipes off the pixels without calculating CRF, excludes non-corner pixels and determines the candidate corners. Then calculates the candidate corners through CRF, obtains the final corner, calculates affine transformation coefficient, and completes the image registration. The experiments show that the algorithm can ensure accuracy, and at the same time, reduce calculation to a great extent, and improve the speed of calculation, so that the automatic image registration can be quickly and effectively achieved, which can help the application in real-time processing.