• Shavet Bhagat
• Sukhwinder Singh

As time passes computer applications are becoming vital part in every field. Due to which it has been used in several applications. Thus detection of object from a video and then tracking of that object is an important task in computer vision. Similarly, the human action recognition has also become as a prominent domain for research work to introduce novel advancements. This study develops a novel human action recognition system on the basis of the Deep Neural Network (DNN). For the purpose of the human recognition system a dataset of images is considered and then the feature extraction is performed from the selected colored and depth image from the dataset. Along with this the segmentation is also applied to the extracted image. After then, the DNN is applied to train the dataset. The reason behind implementing the DNN is to overcome the parametric independent issue of naïve bayes classifier. The simulation of proposed work is done in MATLAB simulation platform and obtained results shows that the proposed work performs the human action recognition with high accuracy.

Human Action Recognition, Feature Extraction, Segmentation, Classifiers, Deep Neural Network.

"New Framework for Analysis of Human Behaviour", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.5, Issue 6, page no.415-421, June-2018, Available :http://www.jetir.org/papers/JETIR1806303.pdf

"New Framework for Analysis of Human Behaviour", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.5, Issue 6, page no. pp415-421, June-2018, Available at : http://www.jetir.org/papers/JETIR1806303.pdf