This paper proposes a panoramic image mosaic method based on feature points matching. This method firstly uses the sift algorithm to extract the image feature points and uses Harris algorithm to optimize the image feature extraction. Then the BBF algorithm based on K-d tree is used to find and determine the initial matching points and complete the coarse matching of the feature points. Then according to the result of image registration, robust RANSAC algorithm is utilized to filter coarse matching feature points and calculate the transformation matrix H. Finally, the gradually fading out fusion algorithm of the weighted average is used in the seamless Mosaic of two images, form a complete panoramic view picture. Experimental results verify the effectiveness of the proposed method and the splicing effect is better.