MSRConvNet: Classification of railway track defects using multi-scale residual convolutional neural network.
Hakan AçikgözDeniz KorkmazPublished in: Eng. Appl. Artif. Intell. (2023)
Keyphrases
- convolutional neural network
- multiscale
- face detection
- pattern classification
- classification method
- decision trees
- feature extraction
- pattern recognition
- preprocessing
- classification systems
- image processing
- support vector machine svm
- classification accuracy
- neural network
- feature space
- multiple scales
- image classification
- model selection
- scale space
- coarse to fine
- classification models
- automatic classification
- classification rules
- classification algorithm
- natural images
- edge detection
- multiresolution
- image sequences
- feature selection
- machine learning