Research On Progressive Transmission Of Multi-resolution Mesh Model Based On Local Feature

With the maturity of 3D scanning technology, three-dimensional models have been widely used in every field of our lives.Three-dimensional models are usually represented by 3D mesh models. With the increase of scanning precision, the 3D mesh model is becoming more and more refined and complex, and the data increases sharply. Meanwhile, people who live in the Internet age browse and share the 3D model more frequently. The speed of 3D model data soared far exceeds the speed of computer hardware processing and the current network transmission ability improved, so the waiting time of getting these three-dimensional models is getting longer. The progressive transmission technology of 3D multi-resolution mesh model provides a good idea for the solution of this problem. However, in the fields of medical three-dimensional imaging, archaeology and three-dimensional image etc, feature regions with edge, dent, tip, folds etc usually need to be observed and processed further.While progressive transmission technology restoring the mesh model layer by layer, users have to spend time waiting for the drawing of large useless regions, which makes the drawing of the ditails of the feature regions is relatively delayed.So users effective waiting time donot become shorter.In order to solve the above problems, in this paper, we propose a progressive transmission algorithm of multi-resolution mesh model based on local feature. On the basis of progressive transmission technology, we give priority to transmit and recover of feature regions to reduce the user’s effective waiting time further. The main research contents of this paper are as follows:(1) A mesh model segmentation algorithm based on vertex local feature is proposed. The region growth algorithm is improved by introducing the concept of local feature of vertices.Through the segmentation, smooth continuous regions with moderate size, distinct different density are obtained, which facilitates subsequent parallel simplification and merging into a complete multi-resolution model. A mesh model simplification algorithm based on vertex local feature is proposed. By improving the classical QEM algorithm, introduce local feature of vertex and local average length of vertex to cost function of edge collapse to delay the feature parts in the edge folding sequence. Not only maintain the feature regions appearance, but also ensure non-feature regions become too great to reflect the overall profile. Finally, the multi-resolution mesh model is generated by the simplified mesh model under the constraint conditions.(2) A progressive transmission algorithm based on the region local feature of multi-resolution mesh model is proposed. Firstly, the multi-resolution mesh model is represented as a base mesh and a series of progressive record information.Then code the base mesh and the progressive record information to reduce the redundant data. Finally, introduce the region local feature and the resolution level, which the progressive recording information belongs to, two constraint factors to sort the progressive record information.Transmit and recover the base mesh first, then the sorted progressive recording information.While the progressive recording information is transmited, recovery it to the base mesh immediately. Thus regions with edge, dent, tip, folds, etc can be obtained priorly, and even regions with more such features can be transmited crossing the resolution level.