Human action in a video based application plays a significant role that alerts the researchers towards recognizing the motion of human. Other video applications also have video content extraction, summarization, and human computer interactions. The existing methods needs manual footnote of pertinent portion of actions of our interest. Recognition of human action can be done authentic without physical commentary of applicable parts of action of any one’s interest. In this paper we try to update the previous reviews on many ways of recognizing Human activities in videos that had different techniques like Hidden Markov model, feature extraction, segmentation etc. the recognition of human activity in applications like visual observation in mobile, human fall detection, video conference, robotics.