Vision based Hand gesture pattern recognition enabled home automation system using Internet of Things

  • R Jayanthi, A Bhuvaneswari, S Rajkumar, K P Rama Prabha
Keywords: Home Automation; SVM; CNN; Hand gestures; Applaud pattern;

Abstract

Controlling electric home appliances and gadgets with the help of switches is difficult for old and disabled people. The main motive of this paper is to help them by automating the control of electrical home appliances using hand gestures and applaud patterns. This paper presents a Support Vector Machine (SVM) and Convolution Neural Network (CNN) based hand gesture recognition system to automate various home devices like lights, fans, etc. This system uses real time image processing for hand gestures recognition by using a simple android-based application and Arduino UNO microcontroller. The captured real-time image by the android application will be processed  using the computer application. The proposed system perform image processing algorithm based on SVM and CNN is used to recognize the input hand gesture. A pre-trained CNN network Resnet50 is used for feature extraction. The CNN feature extraction technique is very powerful, and it outperforms existing hand-crafted feature extraction techniques like HOG,LBP and SURF. The CNN and SVM used together results in more accurate gesture recognition which improves the performance of the system. The system offers an additional applaud pattern recognition module which also can be used by the user for controlling the home appliances.

Published
2021-08-17
How to Cite
S Rajkumar, K P Rama Prabha, R. J. A. B. (2021). Vision based Hand gesture pattern recognition enabled home automation system using Internet of Things . Design Engineering, 8975-8990. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/3459
Section
Articles