Estimating reflectance and shape of objects from a single cartoon-shaded image.
Hideki TodoYasushi YamaguchiPublished in: Comput. Vis. Media (2017)
Keyphrases
- single image
- image data
- object shapes
- visual appearance
- image content
- input image
- surface properties
- image regions
- cast shadows
- object shape
- target object
- image features
- invariant properties
- surface shape
- intrinsic images
- shape properties
- image retrieval
- complex scenes
- multiple objects
- bounding box
- image segmentation
- multiscale
- shape recovery
- surface orientation
- keypoints
- lighting direction
- intensity images
- object classes
- multiple light sources
- moving objects
- lighting conditions
- segmentation algorithm
- point correspondences
- feature points
- occluding contours
- segmentation method
- rigid transformation
- material properties
- shape representation
- viewing direction
- partial occlusion
- curved surfaces
- shape features