The ultimate goal of computer vision is to understand the scene correctly through various steps of acquiring, processing, analyzing and understanding different kinds of information obtained by different kinds of sensors. Human motion analysis, recognition, and understanding is one of the very hottest topics within computer vision. The purpose of this article is to shed some light on the very important subject of human motion understanding, so it can be a good insight for a novice computer vision researcher in this field. The article tries to spot many of the most cited and recent reviews in the field, indicating its wide demanding applications, different taxonomies used in structuring different surveys, various approaches used to solve different stages of the problem, different types of sensors used for data acquisition, some taxonomies used for classifying motions, and various processes involved in motion analysis, followed by a discussion and a conclusion.