Detection of Abnormal Regions on Temporal Subtraction Images based on CNN.
Mitsuaki NagaoHuimin LuHyoungseop KimTakatoshi AokiShoji KidoPublished in: ICBIP (2019)
Keyphrases
- input image
- image features
- region detection
- complex background
- automatically segmented
- region of interest
- anomaly detection
- detection rate
- image data
- test images
- image database
- image regions
- keypoints
- texture and shape features
- image structure
- object recognition
- image retrieval
- image analysis
- ground truth
- image registration
- image classification
- three dimensional
- intensity variations
- image matching
- homogeneous regions
- spatial relationships
- partial occlusion
- pixel intensities
- dynamic textures
- segmented images
- segmentation algorithm
- intensity images
- adjacent regions
- salient contours
- spatial correlation
- feature points
- lighting conditions
- object detection
- detection method
- cellular neural networks
- audio signals
- multiscale
- grey levels
- segmentation method
- detection algorithm
- computer vision