Multiclass defect detection and classification in weld radiographic images using geometric and texture features.
Ioannis K. ValavanisDimitrios KosmopoulosPublished in: Expert Syst. Appl. (2010)
Keyphrases
- multi class
- texture features
- radiographic images
- multiclass classification
- defect detection
- support vector machine
- feature vectors
- multi class classification
- binary classifiers
- feature extraction
- cost sensitive classification
- multiple classes
- multiclass problems
- texture feature extraction
- multiclass support vector machines
- gabor filtering
- cost sensitive
- multi class classifier
- feature set
- texture analysis
- error correcting output codes
- binary classification problems
- feature selection
- binary classification
- gabor filters
- gray level
- low contrast
- image features
- multiclass svm
- support vector
- pairwise
- feature space
- image classification
- imaging modalities
- local binary pattern
- classification accuracy
- training samples
- base classifiers
- automatic identification
- distance measure
- support vector machine svm
- matching algorithm
- pattern recognition
- active learning
- training set
- kernel methods
- image segmentation
- decision trees
- data sets