首页> 外国专利> Improved tfan-based encoding of 3D mesh models composed of triangles

Improved tfan-based encoding of 3D mesh models composed of triangles

机译:改进的基于tfan的由三角形组成的3D网格模型的编码

摘要

Method for encoding a 3D mesh model into a bitstream, comprising: traversing successively each vertex constituting the 3D model; for each vertex traversed: performing a triangle-based fan analysis centred on the current vertex to discover any new triangle(s) not yet discovered when traversing the previous vertices; determining connectivity data, comprising connectivity variables, between vertices based on the newly discovered triangle(s); classifying the values taken by the same connectivity variable into connectivity classes; wherein two independent entropy encoders code the values classified respectively in two separate connectivity classes. Identification data for vertices may include a bitmap. Bitmap and vertex indices may be represented using case predictors of the Scalable Complexity 3D Mesh Compression Triangle Fan standard. Coding cost criteria may be used to select vertex index predictors restricted to positive vertex indices. Predictor flags may be encoded. Sign bits representing sign of residue between current vertex index and numerical value two may be encoded. Degree of vertex may indicate number of times that vertex was identified as a vertex of previously found triangles. Lists of vertices connected to the current vertex may be generated. Number of triangle fan portions may be predicted by half the number of vertices in a list.
机译:将3D网格模型编码为比特流的方法,包括:依次遍历构成3D模型的每个顶点;对于遍历的每个顶点:以当前顶点为中心执行基于三角形的扇形分析,以发现遍历先前顶点时尚未发现的任何新三角形;根据新发现的三角形,确定顶点之间的连通性数据,包括连通性变量;将同一连通性变量采用的值分类为连通性类别;其中,两个独立的熵编码器对分别分类为两个单独的连通性类别的值进行编码。顶点的标识数据可以包括位图。可以使用可伸缩复杂度3D网格压缩三角扇标准的案例预测变量来表示位图和顶点索引。可以使用编码成本标准来选择限于正顶点索引的顶点索引预测变量。预测标志可以被编码。可以对表示当前顶点索引和数值2之间的残差的符号的符号位进行编码。顶点度可以指示该顶点被识别为先前发现的三角形的顶点的次数。可以生成连接到当前顶点的顶点列表。三角形扇形部分的数量可以通过列表中顶点数量的一半来预测。

著录项

  • 公开/公告号GB201610611D0

    专利类型

  • 公开/公告日2016-08-03

    原文格式PDF

  • 申请/专利权人 CANON KABUSHIKI KAISHA;

    申请/专利号GB20160010611

  • 发明设计人

    申请日2016-06-17

  • 分类号

  • 国家 GB

  • 入库时间 2022-08-21 14:08:04

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号