Real-time Sign Language Detection Using Computer Vision and Machine Learning Techniques

Chaitali  Darekar; Sumit  Lawand; Rasika  Mahulkar; Amol  Sagar; Pallavi  Magrulkar

Abstract

Communication plays an important role in our day-to-day lives as it assists people in sharing, and connecting to each other. Despite knowing the significance of the same, some deaf-mute pupils who are not able to communicate freely. These groups of people use sign language for sharing their thoughts. It could be a tough task to understand sign language for a group of humans who are not deaf-mute. We developed a system that makes use of American Sign Language as the solution to the identified problem. The strategy involves implementing the module which detects pre-defined American sign language through hand gestures. For detection of movement of gestures, we use basic hardware components like a camera and OpenCV, media pipelined images and the image is passed to the CNN model for identification followed by word formation. The principal results of using this system will aid deaf-mute humans to share their ideas fruitfully and communicate with normal individuals proficiently and without hesitation.

Abstract

Keywords