Machine learning with openCV to advise robot to move smartly

Project name: Machine learning with openCV to advise robot to move smartly Hardware: (you probably shall have basic hardware) # Raspberry pi. # Simple motors # Pi Camera or similar Software: Python Works: 1. Let me assume that Robot will be placed inside a big working area as shown below. If the ground of area is uniform, and have the same color, same pattern, same structure, …, then probably it does not matter where do we place the robot, the camera will get similar image data from the ground. 2. In real world, some areas inside the working region may be different and we do not want robot to go inside, and similarly, we may not want robot to go outside the working area, as figure shown below. The working area shall have different color, pattern, structure, and so on, compared to the prohibited inside or outside areas. 3. In this regard, when we place the robot inside the working area, we probably shall place the robot at a nice location, and we may allow robot to move a little bit, and let robot to take some images. Robot may define the data from those images as “good” and robot is allowed to pass through. 4. If robot sees the images different from those good images, then robot shall change its moving direction, not pass through the prohibited areas. Below shows two examples. 5. Ideally the working area shall be always same, not change, but let me assume the working area may gradually change its color, pattern, and so on, due to the gradual change of temperature or whatever conditions. If robot can learn by itself, it will be very nice. 6. Thus for this project, basically I want the robot to learn, so it can move smartly.