首页> 外文会议>International conference on 3D Web technology >Coding polygon meshes as compressable ASCII
【24h】

Coding polygon meshes as compressable ASCII

机译:将多边形网格编码为可压缩的ASCII

获取原文

摘要

Because of the convenience of a text-based format 3D content is often published in form of a gzipped file that contains an ASCII description of the scene graph. While compressed image, audio, and video data is kept in seperate binary files, polygonal data is usually included uncompressed into the ASCII description, as there is no widely-accepted standard for compressed polygon meshes.In this paper we show how to incorporate compression of polygonal data into a purely text-based scene graph description. Our scheme codes polygon meshes as ASCII strings that compress well with standard compression schemes such as gzip. The coder is lossless when only the position and texture coordinate {em indices} are coded. If loss is acceptable, positions and texture coordinates can be quantized and delta coded, which reduces the file size further. The gzipped scene graph description files decrease by a factor of two (six) in size when the polygon meshes they contain are coded with the lossless (lossy) ASCII coder.Furthermore we describe in detail a proof-of-concept implementation that uses the Shout3D~cite{shout3d} pure java API---a plugin-less Web3D player that downloads all required java classes on demand. Our prototype is an extremely light-weight implementation of the decoder that can be distributed at minimal additional cost. The size of the compiled decoder class is less than 6KB by itself and less than 3KB if included into a compressed archive of java class files. It makes no use of specific features of the Shout3D API. Hence, our method will work for any scene graph API that allows (a) to extend the node set and (b) to store the scene graph as ASCII.
机译:由于基于文本的格式的便利性,3D内容通常以压缩文件的形式发布,该文件包含场景图的ASCII描述。当压缩的图像,音频和视频数据保存在单独的二进制文件中时,由于没有广泛接受的压缩多边形网格标准,因此多边形数据通常未经压缩就包含在ASCII描述中。在本文中,我们展示了如何合并压缩多边形数据转换成纯粹基于文本的场景图描述。我们的方案将多边形网格编码为ASCII字符串,可以使用标准压缩方案(例如gzip)很好地进行压缩。仅对位置和纹理坐标{em index}进行编码时,编码器是无损的。如果丢失是可以接受的,则可以量化位置和纹理坐标并进行增量编码,从而进一步减小文件大小。使用无损(有损)ASCII编码器编码时,压缩后的场景图描述文件的大小减少了两倍(六),此外,我们还详细描述了使用Shout3D的概念验证实现。 〜cite {shout3d}纯Java API ---无插件的Web3D播放器,可按需下载所有必需的Java类。我们的原型是解码器的极轻量级实现,可以以最小的额外成本进行分发。编译后的解码器类的大小本身小于6KB,如果包含在Java类文件的压缩档案中,则小于3KB。它没有使用Shout3D API的特定功能。因此,我们的方法适用于允许(a)扩展节点集和(b)将场景图存储为ASCII的任何场景图API。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

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