摘要
文章以市内道路场景下采集的图像序列为研究对象,以双目立体视觉中的立体匹配、三维重建、物体分割进行了研究和分析,提出了基于边缘索引的立体匹配算法和基于深度的物体分割算法,最后对算法进行验证,实验结果表明该方法能有效地检测出动态背景下的运动目标。
In this paper, the image sequences collected from the urban road scene are taken as the research object, and thestereo matching, 3D reconstruction and object segmentation in binocular stereo vision are studied and analyzed. A stereo matching algorithm based on edge index and an object segmentation algorithm based on depth are proposed. Finally, the algorithm is verified.The experimental results show that the method can effectively detect moving objects in dynamic background.
出处
《科技创新与应用》
2018年第17期14-15,共2页
Technology Innovation and Application
基金
嘉兴市科技计划项目(编号:2016AY13020)
嘉兴学院2017年度校级重点SRT计划项目
关键词
立体视觉
立体匹配
三维重建
物体分割
stereo vision
stereo matching
3D reconstruction
object segmentation
