i want to creat a project where i have to convert hand gesture i.e sing language into alpahbet....... for that i want to take image from webcam and compare it with the standard imagess
The first step is to generate a Test and Trainer database of Navigation gestures which are to be interpreted using digital camera. The images in both the databases (Test and Trainer) will be stored using.bmp format. These images will be then pre-processed using a normal noise removal method which is Gaussian low pass filter.
These pre-processed images will be then used to obtain feature vectors of each database image. The test database image will be compared with the trainer database images using methods desired such as Thresholding, Magnitude and Phase Gradient Method, Zero Crossing Detection and Principle Component Analysis (PCA).
The appropriate image i.e. the closest match will be selected and the corresponding alphabet/navigation gesture will be displayed on the screen. Simultaneously voice output will also be played for the respective image
now i m confused from where to start so someone plz help................
i have attached block diagram also..