The Research Of Multi-Angle Image Matching Algorithm Based On The Cluster And Mahalanobis Distance

Image matching technology is a hot research field of image analysis,image processing.image information processing.In recent years,with the development of agroforestry techniques.insect identification technology become a hot topic.The study of insect ecology and insect morphology first to study the morphology and structure of insects.The first step of insects image recognition is use the image segmentation technique separated to the recognized matching and the background,image matching regrard the segmented insect image as a template,with the image feature information of the target insect image screened insect image desired,that is another important part of insects image recognition.Therefore,image matching plays an important role in insect image recognition.The insect is an insect image matching technology image recognition indispensable component.SURF algorithm is the research focus for image matching technology in recent years,SURF algorithm Hessian matrix approximation method,and using the integral image greatly reduce the computation time,SURF algorithm is a robust feature with fast image matching method described.So it is widely used in medical image matching,matching remote sensing images and so on.This paper focuses on insect SURF algorithm based image matching algorithm,SURF algorithms optimized for different types of shortcomings in insect Image Matching.The major work and innovations in paper are as follows:(1)SURF insect proposed image matching algorithm based on clustering,The algorithm combines clustering algorithm has a function of detecting the noise point,combined with clustering and SURF algorithm,the feature points of the noise points removed,achieve noise removal function.(2)Paper proposes a multi-angle Mahalanobis distance insects image matching algorithm.SURF-based clustering algorithm can effectively insect image matching,enabling the identification of the image.However,in practice,because of most of the image-capturing different angles,So matching multi-angle image has become a hot topic.However the SURF algorithm for multi-angle image matching is not ideal.Thus,to make the matching algorithm has better affine invariant.The paper use the SURF algorithm combine with the Mahalanobis distance.The results show that SURF insect image matching algorithm based on Clustering either in the absence or in background noise conditions have background noise conditions,Its matching accuracy are much higher than SURF algorithm.Description SURF insect image matching algorithm based on Clustering has a very good application effect in insect image matching.The Mahalanobis distance improved multi-angle insect image matching algorithm improves thetraditional SURF algorithm for matching in multi-angle image matching problem of low efficiency.to match the accuracy rate is much higher than SURF algorithm.Improved algorithm described in this chapter has a very good application effect of multi-angle insect image matching.