Developing a Facial Expressions Recognition Algorithm for Digital Video

Abstract

The automatic detection of facial expressions in digital video is one of the

important research topics at the present time because of its multiple applications imposed by technological development in recent years. Detection accuracy is one of the challenges facing developers of facial expression detection algorithms to achieve the best possible performance. This research focuses on the algorithms of conventional methods, we developed an algorithm based on spatial and temporal stude of video frames to recognize expression change from neutral to one of the another seven expressions, we conducted practical experiments to test its accuracy (experiments showed an accuracy of 98.87% on CK+ dataset), in addition to compare it with algorithm with the same classification.