基于内容理解的单幅静态街景图像深度估计

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (4503 KB) HTML ( ) 输出: BibTeX \| EndNote (RIS)
引用本文:
李乐, 张茂军, 熊志辉, 徐玮. 基于内容理解的单幅静态街景图像深度估计[J]. 机器人, 2011, 33(2): 174-180.. LI Le, ZHANG Maojun, XIONG Zhihui, XU Wei. Depth Estimation from a Single Still Image of Street Scene Based on Content Understanding. ROBOT, 2011, 33(2): 174-180..

摘要提出一种通过分析理解单幅街景图像内的景物构图关系实现图像深度估计的方法.该方法首先对单幅街景图像分块并提取图像块自身的特征以及邻域联合特征，通过机器学习的方法根据图像块的特征识别图像中的各类景物，分析理解街景图像中景物的组成结构；然后，依据小孔成像模型推导出景物的图像坐标和真实深度之间的关系，从而计算出图像内地面区域的深度信息；并根据景物与地面之间的相对位置关系和景物自身的特征变化估计图像内其它景物的深度信息，最终得到整幅街景图像的深度估计结果.实验表明，该方法得到的街景图像深度估计结果能准确反映图像内各个景物在真实世界中的深度分布，在效果上要优于其它的方法.

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS

关键词 ：特征提取, 景物分类, 小孔成像模型, 深度估计

Abstract：A method for depth estimation by understanding how the objects compose the whole scene in a single image of street scene is presented.Firstly,a single image of street scene is segmented into regions.The features of each region and the associated features of its neighbor area are extracted.And the regions are classified as types of object with features of each region by machine learning method,which shows how the image is made up of every object.Then,the depth of ground is estimated by the relationship between coordinate in image and depth in the real world of the same object which is deduced from pin-hole imaging model.And the depth of others in image is estimated by not only the relative position between the objects and ground but also the change of some features in objects.The depth map of image is produced at last. The experiment shows that our algorithm performs better than others and the result of depth estimation reflects the location of each object in the real world exactly.

Key words： feature extraction object classification pin-hole imaging model depth estimation

收稿日期: 2010-06-21 录用日期: 2010-09-30

ZTFLH:

TP391.4

基金资助:国家863计划资助项目（2009AA01Z328）；国家自然科学基金资助项目（60705013，60803101）.

通讯作者: 李乐 E-mail: happylee613@gmail.com

作者简介: 李乐（1983- ），男，博士生.研究领域：多媒体信息系统与虚拟现实技术，机器视觉，嵌入式系统设计．
张茂军（1971- ），男，教授，博士生导师.研究领域：多媒体信息系统与虚拟现实技术，视频图像处理．
熊志辉（1976- ），男，博士，副教授.研究领域：视频图像处理，机器视觉和全向成像技术.

链接本文:

https://robot.sia.cn/CN/ 或 https://robot.sia.cn/CN/Y2011/V33/I2/174

[1] Forsyth D A,Ponce J.Compmer vision:A modem approach[M].Englewood Cliffs,NJ,USA:Prentice Hall,2003.
[2] Hertzmann A,Seitz S M.Example-based photometric stereo:Shape reconstruction with general,varying BRDFs[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2005,27(8):1254-1264.

[3] 张大志,王勇涛,田金文,等.基于单目视觉系统的远距离场景重建算法研究[J].宇航学报,2008,29(1):289-294.Zhang D Z.Wang Y T,Tian J W,et al.Effcient 3D recoustruction using monocular vision[J].Journal of Astronautics,2008,29(1):289-294.
[4] Ens J,Lawrence P.An investigation of methods for determining depth from focus[J].IEEE Transactions on Pattern Analyms and Machine Intelligence,1993,15(2):97-108.

[5] 曾祥尽,黄心汉,吴倩,等.马尔可夫随机场在显微图像散焦深度信息估计中的应用[J].机器人,2008,30(5):416-420.Zeng X J.Huang X H,Wu Q,et al.Application of MRF to depth information estimation of micro image defocus[J].Robot,2008,30(5):416-420.
[6] Saxena A.Chung S H,Ng A Y.3-D depth reconstruction from a single still image[J].Compmer Vision,2008,76(1):53-69.
[7] Hoiem D,Efros A A,Hebert M.Automatic photo popup[C]//32nd International Conference on Computer Graphics and Interactive Techniques.2005:577-584.
[8] Hoiem D,Efros A A,Hebert M.Closing the loop on scene interpretation[C]//IEEE Conference on Computer Vision and Pattern Recognition.Piscataway,NJ.USA:IEEE.2008:1-8.
[9] Hoiem D,Efros A A,Hebert M.Putting objects in perspective[C]//IEEE Conference on Computer Vision and Pattern Recognition.Piscataway,NJ,USA:IEEE,2006:2137-2144.
[10] Felzenszwalb P,Hutteniocher D.Efficient graph-based image segmentation[J].International Journal of Computer Vision,2004,59(2):167-181.

[11] Shotton J,Winn J,Rother C,et al.TextonBoost:Joint appearance,shape and context modeling for multi-class object recognition and segmentation[C]//Europeon Conference on Computer Vision.Piscataway,NJ,USA:IEEE,2006:1-15.
[12] Chang C C,Lin C J.LIBSVM:A library for support Vector machines[EB/OL].(2010-04-01)[2010-05-10].http://www.csie.ntu.tw/～cjlin/libsvm.
[13] Kass M.Wit K A,Terzopoulos D.Snakes:Active contour moodels[J].International Journal of Computer Vision,1988,1[4):321-331.
[14] Havasi L,Sziffmyi T.Extraction of horizontal vanishing line using shapes and statistical error propagation[C]//Photogrammetric Computer Vision Symposium of Internadonal Society for Photogrammetry and Remote Sensing Commission Ⅲ.2006:167-173.
[15] Zafarifar B,Weda H,de With P H N.Horizon detection based on sky-color and edge features[C]//Proceedings of the SPIE,vol.6822.Bellingham,WA,USA:SHE,2008:682220-682220-9.
[16] Duda R,Hart P.Use of the Hough Transformation to Detect Lines and Curves in Pictures[J].Communications of the ACM,1972,15(1):11-15.