Fractional-order differential application and research in pavement cracks image enhancement

Abstract

However, to the best of our knowledge, there are few pavement crack image enhancement researches based on fractional-order differential theory, and almost fewer the study of pavement crack image enhancement, which combines the fractional-order differential with the Prewitt operator. Motivated by the above discussions, we1 INTRODUCTIONThe digital image processing technology, in the application of pavement crack detection, improves the automatization level of pavement crack recognition system based on the camera. In order to analyze the details of pavement crack effectively, there need to identify the enhanced image preprocessing. The existing crack detection algorithm is mainly based on three-dimensional terrain model of pavement crack automatic detection algorithm[1], artificial population algorithm[2], neural network algorithm[3], etc. All above mentioned are improved in a certain extent of regular crack detection algorithm, but for some slight and reticular cracks detection, these methods still cannot achieve the desired effect, and the computation is too large[4]. In digital image processing technology, it is not ideal when dealing with the edge texture details by using one-dimensional wavelet transform tensor product extension of two dimensional. Contourlet transform is due to the presence of the sampling process, which leads to the lack of translation invariance, the result of pavement image enhancement generates pseudo Gibbs distortion and the blurred crack edge.