近年来，对人体动作进行识别和评价正逐渐成为研究和应用热点。动作识别是指对给定的视频或三维动作序列中人的行为动作进行识别，并判断其所属的行为类别。动作评价是指对特定人体动作（如体操、划船、舞蹈等的标准动作）的规范程度或完成质量进行评价，从而辅助人们更好地完成标准动作。动作评价需要基于动作识别来进行。本文以人体动作识别和评价的处理流程为主线，对近年来相关工作进行了全面而系统地梳理，阐述了数据预处理、特征提取、动作分类、动作评价的标准等各个关键环节的研究进展及存在问题，最后对当前研究的难点问题以及未来的发展趋势进行总结与展望。

In recent years, the recognition and evaluation of human actions has gradually become a research and application hotspot. Action recognition refers to recognizing the actions of people in a given video or three-dimensional action sequence, and judging the behavior category it belongs to. Action evaluation refers to the evaluation of the degree of standardization or completion quality of specific human actions (such as gymnastics, rowing, dancing, etc.), so as to assist people to better complete standard actions. Action evaluation needs to be based on action recognition. This paper takes the processing flow of human action recognition and evaluation as the main line, comprehensively and systematically combs the related work in recent years, expounds the research progress and research progress of each key link such as data preprocessing, feature extraction, action classification, and action evaluation standards. There are problems, and finally the difficult problems of current research and future development trends are summarized and prospected.