DeepHAZMAT: Hazardous Materials Sign Detection and Segmentation with Restricted Computational Resources.
Amir SharifiAhmadreza ZibaeiMahdi RezaeiPublished in: CoRR (2020)
Keyphrases
- sign recognition
- cell segmentation
- image segmentation
- automatic detection
- reliable detection
- segmentation method
- false positives
- complex background
- features extraction
- level set
- automated detection
- medical images
- region growing
- fracture detection
- segmentation accuracy
- bounding box
- multiscale
- object segmentation
- edge detection
- segmentation algorithm
- detection method
- energy function
- threshold selection
- landmark detection
- automated analysis
- computed tomography angiography
- boundary detection
- sign language
- false alarms
- gesture recognition
- detection rate
- event detection
- fully automatic
- object detection
- pixel level
- traffic signs
- shape prior
- automatic initialization
- lesion segmentation
- image regions
- face detection
- foreground background separation