Segmentation of Rectangular Objects Lying on an Unknown Background in a Small Preview Scan Image.
Michael GuerzhoyHui ZhouPublished in: CRV (2008)
Keyphrases
- bounding box
- image regions
- background pixels
- segmentation method
- foreground and background
- multiple objects
- foreground background separation
- segmentation algorithm
- image segments
- image segmentation
- figure ground segmentation
- image pixels
- object segmentation
- multiscale
- image data
- image content
- image analysis
- cluttered background
- object shapes
- complex background
- low depth of field
- energy function
- pixel level
- object contours
- edge detection
- cluttered scenes
- image features
- foreground objects
- test images
- segmentation errors
- background clutter
- figure ground
- grey level
- input image
- images of natural scenes
- segmentation accuracy
- segmented images
- salient objects
- color regions
- spatial relations
- target object
- foreground background segmentation
- object regions
- object region
- single image
- textured objects
- graph cuts
- region segmentation
- spatially correlated
- intensity images
- region growing
- outdoor scenes
- pixel wise
- coherent regions
- partial occlusion
- image retrieval
- global thresholding
- segmented regions
- color model
- object shape
- object models
- energy functional
- adaptive thresholding
- polar coordinates
- seed points
- salient regions
- rectangular regions
- background subtraction
- fundamental problem in computer vision
- background model
- superpixels
- shape prior