Open AccessHand Gesture Recognition for Differently Abled People with Message Integration.
Author(s) -
Gaurav Kumar,
G. Vamsi Krishna Reddy
Publication year - 2020
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.d7353.049420
Subject(s) - computer science , region of interest , artificial intelligence , computer vision , convolutional neural network , point (geometry) , locale (computer software) , point of interest , motion (physics) , gesture recognition , subtraction , gesture , background subtraction , pixel , mathematics , arithmetic , geometry , operating system
The objective of this paper is to utilize a webcam to lively track the region of interest (ROI), in particular, the hand locale, in the picture extend and recognize hand motion, we use skin colour discovery and also morphology to delete the unnecessary background information from the picture, and afterward use foundation subtraction to recognize the ROI. Next, to stay away from foundation effects on items or commotion influencing the ROI, we utilize the kernelized connection channels (KCF) calculation to follow the identified ROI. The picture size of the ROI is at that point resized to 28x28 and afterward sent into the profound convolutional neural system (CNN), so as to distinguish various hand signals. Two profound CNN designs are created right now are altered from DenseNet . At that point, the above procedure of following and acknowledgment is rehashed to accomplish a moment impact, and the framework's execution proceeds until the hand is removed from the camera.