说明：  本文主要致力于图像配准和拼接算法的研究，一方面以Harris算法为基础，提出了一种基于圆形邻域增强的角点配准算法，而另一方面则根据图像配准精度需求及庞大图像规模，将图像的拼接算法改进，提出基于尺度不变特征一种的图像拼接算法。
(The thesis focuses on image registration and stitching algorithm, on the one hand to the Harris algorithm, proposed corner registration algorithm based on circular neighborhood enhancements, while the other is based on the accuracy requirements and image registration a large-scale image, the improved image stitching algorithm is proposed based on a scale-invariant feature of image stitching algorithm.)