首页> 外国专利> Image segmentation by means of temporal parallax difference induction

Image segmentation by means of temporal parallax difference induction

机译:通过时间视差差异感应进行图像分割

摘要

An image compositing and compression method based on the creation and processing of parallax differences in motion photography. A parallax scanning MOE lens creates discrete parallax differences in the objects in the recorded scene that are perceived by the viewer as enhanced texture and depth when displayed. Using parallax differences in a captured scene, a computer can detect objects for the purpose of creating image compositing mattes. This method allows matte passes to be filmed on location at the time of principal photography, thereby saving costly additional blue/green stage production shoot days associated with traveling matte techniques. In addition, because the mattes are based on parallax scan differences in the recorded scene and not on a uniform color and luminance process, certain conflicting scene subject colors will not have to be avoided. Also, because the matte scenes are recorded on location, the lighting in each of the various elements matches in the final composited image.
机译:一种基于运动摄影中视差差异的创建和处理的图像合成和压缩方法。视差扫描MOE镜头会在记录场景中的对象中产生离散的视差差,当显示时,观看者会感知为增强的纹理和深度。利用捕获的场景中的视差差异,计算机可以检测对象,以创建图像合成遮罩。这种方法允许在主要摄影时在现场拍摄哑光通行证,从而节省了与旅行哑光技术相关的昂贵的额外蓝/绿舞台制作拍摄日。另外,由于遮罩是基于记录场景中的视差扫描差异,而不是基于统一的颜色和亮度过程,因此不必避免某些冲突的场景主体颜色。另外,由于将遮罩场景记录在位置上,因此各种元素中每个元素的照明在最终的合成图像中都匹配。

著录项

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