This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: HOMEWORK 3 1. Suppose that you want to determine a hand eye transform, that is a transformation from the coordinate system of a mechanical manipulator to that of an electronic camera. This is a problem of exterior orientation. Assume that the camera system can determine the position of the image of a special mark on the gripper of the robot arm. You are free to command the arm to move to any position in its workspace. The transformation, as usual can be broken down into a translation and a rotation. Show that you need at least four calibration points to determine the transformation. 2. In edge based stereo methods, edges in the left image are matched with edges in the right image to obtain disparity measurements. Here we limit ourselves to matching along a single epipolar line, thus reducing the problem to one of matching in one dimension a Suppose there are n edges in each image along an epipolar line. If each edge has a unique match in the other image, how many different mappings are there? Do not include the constraint that edges must be ordered the same way in both images b Now add the constraint that edges must be ordered the same way in both images. If every edge has a unique match, how many different mappings...
View Full Document
- Spring '08
- Rigid Body, Euclidean geometry, Barron, θ