Abstract
Abstract. Accumulating the motion information from a video sequence is one of the highly challenging and significant phase in Human Action Recognition. To achieve this, several classical and compact representations are proposed by the research community with proven applicability. In this paper, we propose a compact Depth Motion Map based representation methodology with hastey striding, consisely accumulating the motion information. We extract Undecimated Dual Tree Complex Wavelet Transform features from the proposed DMM, to form an efficient feature descriptor. We designate a Sequential Extreme Learning Machine for classifying the human action secquences on benchmark datasets, MSR Action 3D dataset and DHA Dataset. We empirically prove the feasability of our method under standard protocols, achieving proven results.
Highlights
In the field of machine vision, Human Action Recognition (HAR) plays a significant role in many applications and it has been one of the active research area for the last two decades
We propose a compact and computationally efficient DMM representation, namely Stridden Depth Motion Map (S-DMM), which hastily generates an accumulated dense structure striding 4 frames of the video sequence at a time
We present experimental results conducted on two standard datasets namely, MSR Action 3D dataset and Depth-included Human Action (DHA) dataset
Summary
In the field of machine vision, Human Action Recognition (HAR) plays a significant role in many applications and it has been one of the active research area for the last two decades. HAR has wide applications and it is an inseparable part of video surveillance, scene understanding, health monitoring, fitness training, gait recognition and human computer interaction. Due to the extensive research works in the domain of image/video processing, action recognition attracted the research community. While dealing with recognition of more complex movements, such as movement of body part either nearer to a camera or away from the camera where there is no texture variation, the distance from body part (object surface) from the camera lens varies significantly. Idea of depth sensors came into reality
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
