MOB LEC3 Cameras and Images

Introduction of computer vision

Computer vision is a field of artificial intelligence (AI) that enables computers derive meaningful information from digital images, videos and other visual inputs obtained by a camera.

s16192210242022

Bucket of photons

1. Photons converted to electrons
2. Shift electrons along row for readout
3. The readout on the device will translate the analog signal to either grayscale images or RGB images

Image Formation

s16250010242022

s16251310242022

pinhole camera, optical centre, focal length

s16260510242022

Stereo Cameras

s16285910242022

s16291210242022

s16292610242022

s16294110242022

Image filtering

s16311810242022

Noise

s16310910242022

Cross-correlation Filter

s16313110242022

s16314610242022

s16322410242022

Image gradient computation

s16330910242022

Summary

• The basics behind imaging and the pinhole camera model.
• 3D points in the world coordinate frame can be projected to 2D points in the image coordinate frame using projective geometry equations.
• These equations rely on the camera intrinsic parameters, as well as its extrinsic location in the world coordinate frame (covered in tutorial).
• Given the geometric transformation between the two cameras of a stereo sensor, and the disparity, you can estimate the 3D location of pixel.
• Cross-Correlation of pixels is the cornerstone of image filtering and can be used to edge detection.

Origin: Dr. Chris Lu (Homepage) Translate + Edit: YangSier (Homepage)