COMPUTER VISION
FROM THE CAMERA TO THE IMAGE
- transform the 3D view in a 2D image
- camera physical parameters
- digitalize an image
- how a camera sensor is made
- lens: effects on the image
- lens distortion
- compute depth using stereo images
- perspective space
- work with perspective projection in perspective space
- homography
- how to calibrate camera
- zhang method
- calibrate a stereo camera system
- warping in computer vision
IMPROVE THE IMAGE BEFORE PROCESSING IT
PROCESS IMAGES
- industrial computer vision pipeline
- binarization
- color based segmentation
- binary morphology
- components labeling
- blob analysis
- finding correspondences
- edges
- canny edge detector
- zero crossing edge detection
- moravec corner detector
- harris corner detector
- shi tomasi corner detector
- how to manage scale invariance
- scale normalized log
- dog detector
- canonical orientation
- sift descriptor
- matching process
DETECTING OBJECTS IN IMAGES
- instance level object detection
- template pattern matching
- shape based matching
- hough transform
- generalized hugh transform
- object detection pipeline