Are there other use cases besides images or data compression?
@datamlistic
4 ай бұрын
Computing the pseudoinverse, eigenfaces, recommendation system etc.
@Kazshmir
4 ай бұрын
It can also be used to find a rigid body transformation between two corresponding point sets. For example if you have a robot arm, in some cases it can be useful to have a 3D camera to tell the robot arm where to go, so you want to know the transformation between camera and robot. One way to get this (rigid body*) transformation is to create two corresponding point sets from the camera and robot, meaning you get multiple common points measured from the cameras perspective and the robots perspective. The SVD can be used to find the transformation matrix when you have these two corresponding point sets. * It's rigid because you are not scaling when transforming from camera to robot coordinates, because both are measured in the same unit: typically meters, or millimeters.
Пікірлер: 4