Fusion features-based broad learning system (FF-BLS) is a classification network with fast and incremental learning. It contains a convolution module and broad learning.
Keras 2
Tensorflow 1.15(<2.0)
Python 3
Opencv 3
BroadLearningSystem.py -the broad learning system
FF-BLStrain.y -main script to extract the convolution feature, start training and testing
Slidewindow-detect.py -Using sliding window and trained FF-BLS model to identify object
BLS-model/ -trained models are saved here
python FF-BLStrain.py #Please modify the dataset load path and model storage path.
Yang Zhang and Ka-Veng Yuen
If there is any question, welcome to discuss here or contact [email protected]
If you use this project in your research, please cite the following paper：
Yang Zhang, & Ka-Veng Yuen* (2021). Crack detection using fusion features-based broad learning system and image processing. Computer-Aided Civil and Infrastructure Engineering, 36(12), 1568-1584.