Fuse-PN: A Novel Architecture for Anomaly Pattern Segmentation in Aerial Agricultural Images.
Shubham InnaniPrasad DutandeBhakti BahetiSanjay N. TalbarUjjwal BaidPublished in: CVPR Workshops (2021)
Keyphrases
- segmentation algorithm
- image analysis
- segmentation method
- test images
- image database
- image data
- segmentation errors
- accurate segmentation
- ground truth
- adaptive thresholding
- grey level
- piece wise
- microscopic images
- edge detection
- segmentation accuracy
- satellite images
- three dimensional
- segmented images
- image segmentation
- unsupervised segmentation
- pixel level
- fully automatic
- image regions
- image features
- input image
- image pixels
- pixel wise
- object recognition
- anomaly detection
- cell nuclei
- image registration
- brain mr images
- image retrieval
- gray level images
- microscope images
- aerial images
- multiple objects
- image segments
- automatically segmented
- image slices
- cardiac magnetic resonance
- object segmentation
- text lines
- gray value
- outdoor scenes
- complex background
- intensity images
- automatic segmentation
- skin lesion
- single image
- multiscale
- fundamental problem in computer vision
- shape prior
- low depth of field