The Research On The Ambiguity Of Marching Cubes In Medical Image Visualization

This paper researched on the ambiguity of Marching Cubes in medial imagevisualization. It analyzed the research background and significance, the research statusat home and abroad, learned about surface rendering and related technologies,researched the basic principles and the main processes of Marching Cubes algorithm.Improved algorithm in this stage of the Marching Cubes algorithm ambiguity problem,some can not completely solve the ambiguity problem, some increase in the complexityof the algorithm, and some increased computational problem, this paper in order toimprove the ambiguity problem, proposed two improved methods: the no-inversesymmetry Marching Cubes reconstruction, the adaptive slicing down split MarchingCubes reconstruction.This paper analyzed that the ambiguity of Marching Cubes can generate cavitationsphenomenon, the reason of this phenomenon is that Marching Cubes considered theinverse symmetry of the cube when generating iso-surface in the cube. Accordingly, weproposed the no-inverse symmetry method, it’s no longer considered the inversesymmetry of the cube, and proposed several line connection rules to generateiso-surface. After analyzed, the no-inverse symmetry method could solve the faceambiguous problems of Marching Cubes. By contrasting the experimental results ofno-inverse symmetry method, traditional algorithm, asymptotic line method and volumerendering algorithm, we found that the capacity to solve the Marching Cubes surfaceambiguities of on-inverse symmetry method is equal to asymptotic line method, and therunning time is better than asymptotic line method.The adaptive slicing down split method divided the ambiguous cube to8equivalentsmall cubes, the normal cubes used original method to rendering iso-surface. Afterdivided, the small cubes equally processed in accordance with this method, until the small cube didn’t have ambiguous problem. Similarly, by programming the adaptiveslicing down split method, Marching Tetrahedral algorithm and volume renderingalgorithm, contrasting the three algorithms’ experimental results, we confirmed that thisimproved method had a good run effects, its experimental results were exactly the sameresults like volume rendering. All of this indicated the feasibility and correctness of themethod. Secondly, the adaptive slicing down split method only split the ambiguouscubes, so its running time didn’t increase too much, the price is worth for generating thecorrect iso-surface. At the same time, the method can improve the precision ofiso-surface at the ambiguities, the experimental results of the visual effect is also closerto the real iso-surface.