Image splicing method based on computer vision

The invention relates to the technical field of image area acquisition, in particular to an image splicing method based on computer vision. The method comprises the following steps of: obtaining the gray level similarity among image blocks according to the gray level distribution condition of the image blocks divided among the images, determining the image block with the best matching relationship, and taking the overlapped image blocks which are not in the matching relationship as adjusting image pairs in each overlapped region which is overlapped for multiple times; adjusting the gray similarity according to the edge distribution of the image blocks in the adjusted image pair to obtain an adjusted similarity index; and according to the adjustment similarity index of the adjustment image pair and the gray level similarity and proportion condition of the non-adjustment image pair, obtaining an overlapping index of an overlapping region and confirming a final region for image splicing. According to the method,