Power point control using hand gesture recognition based on hog feature extraction and k-nn classification
Tejashree Salunke, Sangita Bharkad · 2017
The proposed system is developed using static hand gesture recognition in real-time that facilitates effective and effortless human-computer interaction. This system makes possible the control of Power Point presentation through distance. It is not necessary for the user to control the Power Point presentation through keyboard or mouse or laser pointer. This system does not makes use of traditional methods for hand gesture recognition such as by using hand-gloves, markers, rings, pens or any other devices. The proposed system takes the input data from the portable webcam consisting of four hand gestures. The image captured from the input data is then processed and then histogram of oriented gradients features is extracted from it. The processed image is then compared with the database of gesture images. Image is compared and recognized using K-nearest neighbor algorithm. The recognized image is then used to control the Slide-Show Presentation. The system is tested in different kinds of light sources - dull, medium, and bright. Gesture images are properly detected when the background consists of bright light.