Specific Motion Pattern Detection: State-of-the-Art and Challenges

Abstract

Detecting motion patterns in video constitutes a field which requires complex computations as processing of moving frame (video) is to be done. The field requires detection of patterns in motion by evaluating each frame pixel by pixel. Since we are mostly interested in detecting only specific patterns, it is possible that patterns that are not specific could be discarded. Applications of motion pattern detection cut across several verticals including surveillance, storage optimization, video compression, tracking of objects and determining the general direction of motion of multiple objects. In the present work, we have explored the state-of-the-art techniques and algorithms for detection of motion patterns, along with the associated challenges.