Damage Classification for Masonry Historic Structures Using Convolutional Neural Networks Based on Still Images.
Niannian WangQingan ZhaoShengyuan LiXuefeng ZhaoPeng ZhaoPublished in: Comput. Aided Civ. Infrastructure Eng. (2018)
Keyphrases
- image classification
- image data
- image database
- input image
- object recognition
- handwritten digits
- ground truth
- convolutional neural networks
- three dimensional
- classification method
- image collections
- image structure
- classification accuracy
- image analysis
- pattern recognition
- pigmented skin lesions
- decision trees
- feature space
- image registration
- edge detection
- preprocessing stage
- thematic mapper
- machine learning
- test images
- convolutional network
- image regions
- image features
- support vector machine svm
- image matching
- visual features
- region of interest
- feature points
- supervised learning
- cultural heritage
- extracted features
- support vector machine
- high resolution
- support vector
- feature extraction
- image processing