SIFT算法。成像匹配的核心问题是将同一目标在不同时间、不同分辨率、不同光照、不同位姿情况下所成的像相对应。传统的匹配算法往往是直接提取角点或边缘，对环境的适应能力较差，急需提出一种鲁棒性强、能够适应不同光照、不同位姿等情况下能够有效识别目标的方法。(SIFT.The core problem of imaging matching is to correspond to the image formed by the same target at different times, different resolutions, different illuminations, and different poses. The traditional matching algorithm often extracts corners or edges directly, and has poor adaptability to the environment. Therefore, it is urgent to propose a method that is robust, adaptable to different illuminations, different poses, etc., and can effectively identify the target.)