Object avoidance bot using image acquisition

I want to create a bot which processes a image and avoids obstacles on processing the image.