Normalized Joint Mutual Information Measure for Image Segmentation Evaluation with Multiple Ground-Truth Images.
Xue BaiYibiao ZhaoYaping HuangSiwei LuoPublished in: CAIP (1) (2011)
Keyphrases
- ground truth
- mutual information
- similarity measure
- image registration
- segmented images
- image segmentation
- image segmentation algorithms
- image analysis
- input image
- test images
- gold standard
- image database
- segmentation method
- quantitative evaluation
- image data
- ground truth data
- segmentation algorithm
- three dimensional
- normalized cross correlation
- information theoretic
- object recognition
- image features
- segmenting images
- human subjects
- high quality
- unsupervised segmentation
- criterion function
- markov random field model
- information theoretic measures
- multimodal image registration
- information gain
- maximization of mutual information
- normalized correlation
- image retrieval
- sum of squared differences
- adaptive thresholding
- image classification
- unsupervised image segmentation
- image regions
- evaluation measures
- expectation maximization
- manually labeled
- active contours
- registration accuracy
- gray level
- medical images
- edge detection
- deformable models
- multiscale
- region growing
- computer vision