Image-only place recognition based on regional aggregating ConvNet features for underground parking lots.
Xianglong WangXinyuan ZhuZhongzhen YanZhiwei YeJiangyi DuFeng GuoZhigang XuChun LiuCailu MaoPublished in: Vis. Comput. (2024)
Keyphrases
- image features
- low level
- extracting features
- image data
- extracted features
- input image
- test images
- image analysis
- image pixels
- original images
- single image
- image description
- invariant features
- image regions
- feature representation
- image content
- grey level
- visual appearance
- spatial information
- image segmentation
- matching process
- feature detectors
- spatial distribution
- keypoints
- image retrieval
- geometrical features
- feature extractor
- image collections
- image similarity
- feature matching
- sample images
- salient features
- image set
- image classification
- feature extraction
- high resolution
- static images
- object recognition
- edge detection
- color features
- image representation
- individual features
- multiscale
- spatial relationships
- similarity measure
- textural features
- feature space
- feature vectors
- feature values
- image database
- sift descriptors
- target object
- natural images
- segmentation algorithm
- region of interest
- color histogram
- feature descriptors