Research On Skeleton Optimization Algorithm Based On Medial Axis

The medial axis is an abstract representation of image,which can be regarded as a refined structure of image.It retains the original topological structure and feature information of image,and is easy for computer to save and transfer the structural feature information of image.In order to solve the problem of too many redundant branches of The medial axis and the lack of conciseness in the whole,many scholars have carried out relevant research.Among them,the Bayesian model method proposed by Feldman.applies Bayesian rule to skeleton generation,and prunes The medial axis through optimization algorithm to obtain the optimal skeleton.In this paper,the idea of Feldman Bayesian model is applied to the research of skeleton optimization.Under the premise of controlling image reconstruction error,balancing algorithm is established by using the reconstruction error and skeleton conciseness as control parameters for optimize the skeleton branches contained in The medial axis and achieve the balance and unification of reconstruction accuracy and skeleton simplification.The whole algorithm adopts the reverse pruning method.The skeleton principal axis is established by The medial axis and the branching level is set,Starting from the skeleton spindle,the optimal approximate skeleton of the image is formed by adding the optimal medial axis branches into the skeleton spindle by using the control parameters and the balance algorithm to optimize the medial axis branches according to the level.The object of this experiment is binary image.The skeleton is reconstructed by morphological knowledge.When analyzing the experimental results in detail,two skeleton pruning algorithms are selected to compare the results.The experimental results show that the proposed algorithm is effective and robust.The complexity of the algorithm is low and the operation speed is fast.Under the control of image reconstruction error,the skeleton obtained is more concise.