The Method of Damaged Pavement Image Splicing Based on Image Pyramid and FCM Clustering

There are usually large noise and anomalistic distribution of characteristic points in the damaged pavement images, which bring great difficulty to splice. In view of the characters, this paper forms one kind of splicing method of the pavement images which is based on image Pyramid, the Harris vertex examination and FCM clustering. Firstly, it forms several groups of the structure of the criterion spaces carrying on the transformation of the multi-criteria and multi-resolution Pyramid. Then it takes this stratification structure as the foundation, and carries on the Harris vertex examination to each structure. Thus it makes the characters carry on the FCM clustering and realize splicing of neighboring damaged pavement surface images.