tailieunhanh - Lecture Digital image processing - Lecture 8: Camera Calibration and Stereo Imaging

After studying this chapter you will be able to understand: described how memory stores data, instructions, and information, and discussed the sequence of operations that occur when a computer executes an instruction. The chapter included a comparison of various microprocessors on the market today. | Digital Image Processing CCS331 Camera Calibration and Stereo Imaging 1 Summery of previous lecture Inverse perspective transformation imaging geometry where the world coordinate system and the camera coordinate system are perfectly aligned what are the transformations steps involved in a generalized imaging setup where the world coordinate system and the camera coordinate system not aligned Illustrate the concept with the help of an example 2 Todays lecture Calibration of the camera Extracting the 3D point from 2 images which is also known as stereo images Stereo correspondence problem complexity 3 Not perfectly aligned camera The camera can be a pan of angle theta, it can also be given a tilt by an angle alpha. 4 For a given world point W, what will be the corresponding image point C? we will need to find a set of transformations to bring the camera coordinate system and the world coordinate system in perfectly aligned. Then apply the perspective transformation to the transformed 3D world points to the corresponding image coordinates 5 Transformation steps we have 4 different transformation steps which are to be applied one after another and these transformation steps will give you the transformed coordinate of the 3D world point W. 6 Displacement If the camera center is displaced by vector W0; so all the world coordinate points, will be displaced by the vector which is negative of W0 7 panning the camera Pan the camera by angle theta and this panning is done along the Z axis. 8 Tilt the camera we have to find out what is the corresponding transformation matrix for this tilt operation which has to applied to all the 3D points. 9 combined into a single rotation matrix 10 camera center transformation Then final transformation is the camera center from the Gimbal center by a vector R 11 All transformation one after another 12 13 Calibrate camera The transformation which is involved Transformation matrices are of dimension 4 by 4; This matrix A will also be a 4 . | Digital Image Processing CCS331 Camera Calibration and Stereo Imaging 1 Summery of previous lecture Inverse perspective transformation imaging geometry where the world coordinate system and the camera coordinate system are perfectly aligned what are the transformations steps involved in a generalized imaging setup where the world coordinate system and the camera coordinate system not aligned Illustrate the concept with the help of an example 2 Todays lecture Calibration of the camera Extracting the 3D point from 2 images which is also known as stereo images Stereo correspondence problem complexity 3 Not perfectly aligned camera The camera can be a pan of angle theta, it can also be given a tilt by an angle alpha. 4 For a given world point W, what will be the corresponding image point C? we will need to find a set of transformations to bring the camera coordinate system and the world coordinate system in perfectly aligned. Then apply the perspective transformation to the transformed