Features extraction methods for object detection with Visual Cameras and path planning?

edit

Hi

I would like to use the Vision Cameras ( like Kinect or any other Vision System) to identify the position of an object (like car door aircraft door) and navigate the robot there (to the door). So I was thinking to use a features extraction methods for object recognition (detection, identifying) and than some navigation(path planning) method to navigate to that object

Im not familiar with feature extraction methods using Kinect or any other Vision Cameras so would like to ask for some help to hit the ground. Any tutorials or packages or methods that can help and start using it and build my algorithms from there?

Also any path planning and navigation methods using Visual Cameras?

Thanks