Human Computer Interaction System Base On Action Recognition

The history of the development of human-computer interaction technology, has beenmore than fifty years in recent years, the implementation of human-computer interactionbecomes one of the key directions of scholars at home and abroad by the gesturerecognition technology. Action recognition method based on active optical sensor trackingis the development of the most mature, the most widely technical application type. With ahigh research significance and practical value. Commonly used method is the depth of ascene graph analysis of data acquired by a sensor to judge, recognition and action ofhuman body parts, so as to realize the action recognition based on human computerinteraction.This artical analyses the technical characteristics and application scope of thetechnique principle, human-computer interaction, and focus on the research in the Kinectenvironment, action recognition method based on a depth map in human-computerinteraction technology realization and application specific, and obtained the followingresearch results:(1) Proposed the separation algorithm for human and environmental background in theimage by computing the depth, the depth value, comparison of the results, using the regiongrowing algorithm will be the target of the human body and the background effectiveseparation(2) Proposed the separation algorithm for human and environmental background in theimage by computing the depth, the depth value, comparison of the results, using the regiongrowing algorithm will be the target of the human body and the background effectiveseparation(3) Template matching algorithm to identify a single motion trajectory based on parts ofthe human body, and the algorithm was verified by experiment, the efficiency rate was91%. (4) Development of an automatic presentation system based on the application of the depthdata action recognition.