作者: JaChing Chou , W.A. O'Neill , H.D. Cheng
DOI: 10.1109/ICSMC.1994.399871
关键词: Image segmentation 、 Entropy (information theory) 、 Backpropagation 、 Computer science 、 Invariant (mathematics) 、 Artificial neural network 、 Contextual image classification 、 Artificial intelligence 、 Entropy (energy dispersal) 、 Pattern recognition
摘要: A novel approach of applying moment invariants and neural networks to analyze pavement images is presented in this paper. By calculating from different types distress, features are obtained. Then a backpropagation network used classify these features. This illustrated using randomly selected sample video real cracks. Based on samples, the feasibility crack proven. >