In view of the fact that feature point extraction speed is slow and that its matching accuracy is not high in the current image stitching process,an optimized image stitching algorithm is proposed.Firstly,the images to be spliced are carried by the downsampling process; then extract the feature points on the half image area and use the Sequential Similarity Detection Algorithm (SSDA) to purify the feature points; finally,stitch the images.Stitching results show that compared with the traditional image matching method,the new optimization algorithm greatly reduces calculated data and has obvious advantages in image stitching time.