Image Pre-Processing based Sign Language Recognition using Convolutional Neural Network
T. Tamilvizhi, R Surendran, Geetha Rani K
Abstract
The domain of the project is Image pre-processing. One of the most commonly used ways of expressing oneself to others is communication. Communication is a virtue of Life. Still a correspondence between a visually challenged individual and a Hearing challenged and Speaking challenged individual is a nightmare. A Visually challenged person can communicate to normal people easily by speech, whereas the Hearing challenged and speaking challenged people can communicate using Sign language. But the communication among them is not possible in this advancing technology. Visually challenged person cannot understand the Sign language, which Hearing challenged and Speaking challenged people use. And Hearing challenged and speaking challenged people cannot hear what visually challenged person says. Hence, we propose a system for communication between a visually challenged, Hearing challenged and Speaking challenged people using HSV (Hue Saturation Value). Herein we use a Sign language to detect what Hearing challenged and speaking challenged people say to visually challenged person and vice versa. To begin with, we make a classifier model utilizing the signs utilizing the Keras execution of convolutional neural organization utilizing python. Then, at that point, we measure another continuous framework which utilized skin division to discover the Region of Interest in the casing which shows the jumping box. The classifier observed to be improving with various foundations and hence the point of the picture caught.