Convolutional Neural Networks for Steel Surface Defect Detection from Photometric Stereo Images.
Daniel SoukupReinhold Huber-MörkPublished in: ISVC (1) (2014)
Keyphrases
- stereo images
- defect detection
- convolutional neural networks
- textured surfaces
- stereo matching
- stereo method
- image pairs
- stereo pair
- epipolar lines
- stereo vision
- stereo correspondence
- stereo reconstruction
- dense disparity map
- feature extraction
- planar surfaces
- stereo camera
- stereo matching algorithm
- specular reflection
- disparity map
- surface reconstruction
- stereo algorithm
- automated visual inspection
- real scenes
- object surface
- depth discontinuities
- image rectification
- smooth surfaces
- photometric stereo
- d objects
- ground truth
- matching cost
- reflectance properties
- image processing
- three dimensional
- viewpoint
- dynamic programming
- depth map
- real time