Human Body Posture Recognition Using Artificial Neural Networks
Manu Bali1, Devendran V2
1Manu Bali, Lovely Professional University, Punjab, India.
2Devendran V, Lovely Professional University, Punjab, India.
Manuscript received on 05 April 2019 | Revised Manuscript received on 14 April 2019 | Manuscript Published on 24 May 2019 | PP: 230-234 | Volume-8 Issue-6S3 April 2019 | Retrieval Number: F10470486S319/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Activity acknowledgment has turned into a critical theme in recognizing the passionate action with numerous crucial applications. To abstract the concept, human detection is the very initial step to be performed in each field like surveillances, abnormal behaviour detection, crowd analysis and etc. In human association, understanding human practices is a testing issue in this day and age. Various poses of the body will be the challenging always for activity recognition. Hence, orientation / pose of the body detection become a vital one. In this work, Head is used as the key for body-pose recognition. Here, we present a framework, which is capable of finding the posture orientation of a human body in context of head position. Extracted features are used to train Artificial Neural Networks to obtain the Head Position. Our work consists of following steps: (i) Human Detection (ii) Upper Body / Lower Body Detection (iii) Head Detection (iv) Human Pose Recognition. This work is implemented in Matlab Platform. Datasets are taken randomly from Google Images.
Keywords: Boundary Extraction, Upper and Lower Body Segmentation, Body-Pose Recognition.
Scope of the Article: Artificial Intelligence and Machine Learning