2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA)

1 / 2018-06-16 20:46:27

Image Registration Algorithm Based on Super pixel Segmentation and SURF Feature Points

Image registration,SURF feature point detection,information entropy, semantic segmentation

In the current image registration technology, the accuracy of feature point detection and matching feature point pair is relatively low. Based on the analysis of image registration based on SURF feature point detection and information entropy, a SURF feature based on semantic partitioned information entropy Point image registration algorithm. Firstly, super-pixels are used to segment the image region, and the information entropy of each image region is calculated. The problem of the SURF operator distributing more densely when extracting feature points is improved. The use of the amount of information to exclude part of the feature points redundant points, reducing the number of feature points. Experimental results show that the improved algorithm can improve the accuracy of image registration and effectively improve the quality of registration.