A multimodal approach to time-invariant scene retrieval from single overhead image.
Jian YaoZhongfei (Mark) ZhangLijun YinPublished in: Multimedia Information Retrieval (2004)
Keyphrases
- input image
- single image
- image retrieval
- complex scenes
- image content
- image description
- visual appearance
- image data
- image regions
- image set
- scene understanding
- image features
- d scene
- piecewise planar
- scene classification
- three dimensional
- reference images
- image segmentation
- image representation
- image database
- similarity matching
- high resolution
- scene images
- uncalibrated cameras
- imaging process
- image frames
- image collections
- spatial relations
- image classification
- spatial information
- scene matching
- multiscale
- location and orientation
- relevance feedback
- calibrated cameras
- segmentation method
- geometric information
- outdoor scenes
- information retrieval
- ground plane
- visual data
- image matching
- geometric constraints
- scene geometry
- lighting conditions
- vanishing points
- camera images
- low level
- multiple objects
- image sequences