Analysis And Research Of Medial Axis Computation For Planar Domain Shapes

With the rapid development of computer technology, the fields of computer graphics applications are increasingly being used, so that important researches in computer graphics as a graphic image processing techniques are more and more attention. In this research field, how well the shape of an object that is a fundamental issue, the current solution to this problem the most widely used method is the medial axis, said method in computer animation, computer-aided design/manufacturing, surface fitting, computer vision, robot path planning, finite element mesh, pattern recognition, image analysis, geographic information systems has already been applied, and rapidly extended to other areas. Thus, in the axial expansion of application fields, while the medial axis extraction algorithm on increasingly high demand for quality, how fast and accurate extraction of the medial axis of a medial axis of the object of research and development priorities.In this paper, we take the extraction of planar domain's medial axial as the research object. Firstly, we introduce the background and significance of the medial axis of the medial axis of extraction of domestic and international status. Study based on the maximum disk and grass-fire methods are based on the two extraction model, and comparative analysis of advantages and disadvantages of the two models, leads the research trend in this research.Secondly, the medial axis through the introduction of existing extraction algorithm, the current mainstream in the field were analyzed and compared, pointing out that the advantages and disadvantages of related algorithms, and the polygon axis extraction and refinement based on morphology is improved, mainly axial deformation of the tedious process of extracting a simplified and improved the morphological thinning algorithm, so that continuous improvement is more consistent with human visionly.Finally, this study will focus on the planar domain's medial axis extraction algorithm, but also that using the planar domain boundary curves and the precise geometry of the medial axis to strike the medial axis. Extraction of the geometric axis from the principle, described in detail how to use the knowledge of differential geometry, accompanied by extraction to build the medial axis equidistant from the intersection of the relational model and process model to study the boundary curves and their corresponding medial axis of the differential geometry relationship. On this basis, through the boundary line of the axis of curvature information to find the initial point, the use of tracking and iterative algorithm to calculate the next point, which can be more precise and efficient, and finally get the axis of the original object by interpolating. At this point, we can reconstruct the original object through the medial axis of the cluster transformation model and the envelope of plane curve theory.