首页> 外国专利> Method and apparatus for automatic segmentation of a foreground object in an image

Method and apparatus for automatic segmentation of a foreground object in an image

机译:用于自动分割图像中的前景对象的方法和设备

摘要

The present invention relates to a method for segmenting a foreground object in an image, comprising the steps of: providing a digital scene (1) with a to be segmented foreground object and a background, providing a the background representing the digital reference image (2), generating a the pixel by pixel difference between the image and the reference image representing the difference image (3), means of pixel-by-pixel comparing the difference image (3) with a certain threshold value, allocation of each pixel of the scene (1) either to the background or the foreground object on the basis of the threshold comparison and generating a result of containing the segmented foreground object image (7). The present invention is characterized in that the threshold value is determined that, at first in the differential image (3) a spatial estimate of the pixels to be assigned to the foreground object due to the outer contours and the threshold value then carried out by means of one or more properties of the set of the remaining pixels of the difference image (3) is determined.
机译:本发明涉及一种用于在图像中分割前景对象的方法,包括以下步骤:提供具有要分割的前景对象和背景的数字场景(1),提供表示数字参考图像的背景(2)。 ),在图像和代表差异图像的参考图像(3)之间生成一个像素一个像素的差异,将差异图像(3)与某个阈值进行逐像素比较的方法,分配像素的每个像素在阈值比较的基础上将场景(1)与背景或前景对象进行比较,并生成包含分割后的前景对象图像(7)的结果。本发明的特征在于,确定阈值,首先,在差分图像(3)中,由于外轮廓而对要分配给前景对象的像素进行空间估计,然后通过确定差异图像(3)的其余像素组的一个或多个属性中的一个。

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