Signtium: Real Time Dynamic Hand Gesture Recognition System

The project Signtium is a unique approach for hand gesture recognition. It uses a mediocre web camera and Intel Pentium based personal computer to recognize gestures. The text presents the details of the line of action for making the computer acquire the hand, distinguish between the direction of movements of hand and shapes a hand makes by fingers. The algorithm implemented employs the digital image processing techniques and computer surmises to gestures by artificial intelligence using neural network. The project features mainly three applications: Firstly it controls the windows programs that include games, media player and other applications. Secondly, it takes its approach towards robotics and last but not least it controls a household practice that is controlling appliances. All the applications were a part of demonstration. Signtium is implemented in C++ using OpenCV libraries. The core of this report enumerates some techniques and describes the mechanism of working of the project Signtium.