Hand gestures would be an intuitive and ideal way of interacting with computers, and would require little more than a digital camera. One could, for example, wave downwards to scroll through pages. As a step towards this goal, we attempted to create a program that would recognize and classify hand gestures from a "first person" perspective in a static environment. By isolating the most obvious features of the hand - the fingers - we determine what gesture is being formed by the hand. Sample data were taken of a hand in various positions, orientations, and gestures. While our program was able to identify hand gestures from four classes - open hand, closed fist, pointing, and gripping - it was not accurate enough to be practical in a natural environment.