Reconstruction of an all-in-focus image by region-adaptive fusion of limited depth-of-field images.
Jae Young LeeSung Oh ParkRae-Hong ParkPublished in: ICCE (2016)
Keyphrases
- input image
- image regions
- source images
- image data
- reconstruction process
- region of interest
- image pixels
- test images
- image collections
- image features
- segmented images
- image structure
- grey level
- image retrieval
- depth map
- original images
- computerized tomography
- back projection
- image set
- high resolution images
- image analysis
- image fusion
- adjacent regions
- image classification
- homogeneous regions
- fused image
- depth cues
- defocus blur
- pixel values
- image content
- high resolution
- gradient information
- piecewise planar
- three dimensional
- segmentation method
- segmentation algorithm
- reconstruction method
- imaging technology
- dense depth map
- pixel level
- image matching
- edge detection
- segmented regions
- region segmentation
- image reconstruction
- feature points
- intensity distribution
- foreground and background
- higher resolution
- gray level
- depth information
- image database
- stereo matching
- lighting conditions
- image segmentation
- image quality
- region boundaries
- raster scan
- band limited
- multiscale
- image processing
- image registration
- maximum a posteriori probability
- depth from defocus
- face images
- matching score
- target object
- bounding box