Hand Motion Analysis using CNN

Harsh Raj*,Aditya K Shetty M,Aditya Duggal,Sreekanth Uppara,Srividya M S

doi:10.35940/ijsce.f3409.059620

Harsh Raj*, Aditya K Shetty M + Show 3 more

Open Access

https://doi.org/10.35940/ijsce.f3409.059620

Copy DOI

Abstract

-Hand motion detection and gesture recognition research has attracted large interest due to its wide range of applications in the field of Human computer interaction such as sign language recognition, 3D printing, virtual reality. There have been several approaches to create a robust algorithm to ease human computer interaction and perform in unfavourable environments. The real time recognition and learning of the model are big challenges. In this work, we use Convolutional Neural Network architecture to detect and classify hand motions, the region of interest of the image is passed through the neural network for the hand motion analysis and detection. Our system has achieved testing accuracy of 98%.

Highlights

Human computer interaction has expanded covering most parts of the forms of modern information technology design available.Among the several human computer interaction approaches, Hand motion detection and analysis is one of them. It can be used as a way to ease the interaction between human and computer.In the existing system, hand motion detection and analysis are affected by the working conditions. in this work we try to overcome the cluttered environment and produce a satisfactory result and try to ease the Human computer interaction
Background subtraction, binary thresholding, Grayscaling along with Gaussian filtering are used for converting images received from the camera into Binary Images fit to be fed to the Deep Convolutional Neural Network
This work extracts the circumstances and issues in analysis of hand motions using an application of artificial neural networks

Summary

Introduction

Human computer interaction has expanded covering most parts of the forms of modern information technology design available.Among the several human computer interaction approaches, Hand motion detection and analysis is one of them. It can be used as a way to ease the interaction between human and computer.In the existing system, hand motion detection and analysis are affected by the working conditions. in this work we try to overcome the cluttered environment and produce a satisfactory result and try to ease the Human computer interaction. Hand motion detection works properly when the factors are favorable to the camera. An effective real time hand motion detection system is developed which works overcoming the challenges faced in the earlier systems

Methods

Results

Conclusion