一种基于MRF的单幅图像数据的三维重构方法研究

李蓉; 邓春健; 邹昆

doi:10.3788/YJYXS20163103.0301

您当前的位置：

首页 >

文章列表页 >

一种基于MRF的单幅图像数据的三维重构方法研究

显示与图像 | 更新时间：2020-08-12

- 一种基于MRF的单幅图像数据的三维重构方法研究
- 3 D reconstruction method based on single image data by MRF
- 液晶与显示 2016年31卷第3期页码：301-309
- 作者机构：
  
  1. 电子科技大学中山学院,广东中山,528402
  2. 电子科技大学计算机科学与工程学院,四川成都,611731
- 作者简介：
  
  [ "李蓉(1982-),女,四川绵阳人,硕士,讲师,主要从事机器视觉、多媒体技术等方面的研究。E-mail:windylee007@126.com" ]
  [ "邓春健(1980-),男,广东乐昌人,博士,教授,硕士生导师,主要从事嵌入式系统、智能设备等方面的研究" ]
  [ "邹昆(1981-),男,湖北郧西人,博士,副教授,主要从事图形图像处理等方面的研究。" ]
- 基金信息：
  
  国家自然科学基金项目(No.61302115)();广东省自然科学基金(No.S2013010015764)();广东省高等学校优秀青年教师培养计划项目(No.Yq20
- DOI：10.3788/YJYXS20163103.0301
  中图分类号： TP334.8;TP391.9
- 收稿日期：2015-09-17，
  
  修回日期：2015-12-01，
  
  纸质出版日期：2016-03-05
- 稿件说明：
移动端阅览
聂鹏, 孔玲君,. 空间细分插值的多色打印机分色算法[J]. 液晶与显示, 2016,31(3): 317-323

NIE Peng, KONG Ling-jun,. Color separation algorithm of multi-color printer based on space subdivided with interpolation[J]. , 2016,31(3): 317-323
聂鹏, 孔玲君,. 空间细分插值的多色打印机分色算法[J]. 液晶与显示, 2016,31(3): 317-323 DOI： 10.3788/YJYXS20163103.0301.

NIE Peng, KONG Ling-jun,. Color separation algorithm of multi-color printer based on space subdivided with interpolation[J]. , 2016,31(3): 317-323 DOI： 10.3788/YJYXS20163103.0301.

摘要

综合分析了常见的基于图像的三维重构方法的优缺点

提出了一种基于单张图像

采用马尔科夫随机场(MRF)推断3D位置和方向的3D重构方法。该算法首先将图片分割成多个小的区域(超像素块)

并假定空间场景由许多很小的平面组成

超像素块与平面相互对应

对图像中每个超像素块求取出一组特征向量(纹理、颜色等)

使用MRF模型化平面参数之间、超像素特征向量与平面参数之间的关系

采用监督学习的方式求取相关参数

求解MRF模型

并根据平面参数进行场景重建。这种算法不需对场景结构做明确的假定

因此较之以前的方法可以获得更多3D结构细节信息。用该方法对200张图片样本进行验算

样本中有60%实现了较为准确的3D重构。

Abstract

The mainstream methods of image-based 3D reconstruction are overviewed. The advantages and disadvantages of these methods are analyzed. A new 3D-reconstruction method is put forward; it is based upon a single image of an unstructured environment. For each small homogeneous patch in the image

a Markov Random Field (MRF) is used to infer a set of parameters that contain both the 3D location and 3D orientation of the patch. The space is assumed being made up of many tiny planes

and each plane corresponds with a super-pixel. For each super-pixel in the image

a series of characteristic vectors such as color and texture are calculated. The relations between different planar parameters

as well as between super-pixel vectors and planar parameters are probed by using the MRF model. The MRF is trained via supervised learning. The scene is reconstructed based on the planar parameters. This algorithm does not require explicit assumptions about the structure of the scene

and compared with previous methods it enable us to gain more detailed information about the 3D structure. Using this approach

we have conducted checking computations for 200 pictures

among which 60% have been correctly 3D reconstructed.

关键词

Keywords

references

佟帅,徐晓刚,易成涛,等.基于视觉的三维重建技术综述[J].计算机应用研究,2011,28(7):2411-2417. TONG S, XU X G, YI C T, et al. Overview on vision-based 3D reconstruction[J]. Application Research of Computers, 2011, 28(7):2411-2417. (in Chinese)

王欣,袁坤,于晓,等.基于运动恢复的双目视觉三维重建系统设计[J].光学精密工程,2014,22(5):1379-1387. WANG X, YUAN K, YU X, et al. Design of binocular vision 3D reconstruction system based on motion recovery[J]. Optics and Precision Engineering, 2014, 22(5):1379-1387. (in Chinese)

邹虹,曾鑫,温鑫.基于两视点视频融合技术的裸眼3D显示的研究[J].液晶与显示,2014,29(5):824-829. ZOU H, ZENG X, WEN X. Glasses-free 3D display based on two-view video synthesize technology[J]. Chinese Journal of Liquid Crystals and Displays, 2014, 29(5):824-829. (in Chinese)

姜翰青,章国锋,董子龙,等.基于图像序列的交互式快速建模系统[J].计算机辅助设计图形与图形学学报,2008,20(9):1196-1203. JIANG H Q, ZHANG G F, DONG Z L, et al. Fast interactive scene modeling from an image sequence[J]. Journal of Computer-Aided Design & Computer Graphics, 2008, 20(9):1196-1203. (in Chinese)

段福庆,吴福朝,胡占义.基于平行性约束的摄像机标定与3D重构[J].软件学报,2007,18(6):1350-1360. DUAN F Q, WU F C, HU Z Y. Camera calibration and 3D reconstruction using parallelism constraint[J]. Journal of Software, 2007, 18(6):1350-1360. (in Chinese)

HORN B K P.Shape from shading:A method for obtaining the shape of a smooth opaque object from one view[D]. Massachusetts:Massachusetts Institute of Technology, 1970.

吴凤和,张晓峰,施法中.基于单幅图像数据的三维重构方法研究[J].中国机械工程,2007,18(17):2071-2075. WU F H, ZHANG X F, SHI F Z. Research on 3D reconstruction method based on single image data[J]. China Mechanical Engineering, 2007, 18(17):2071-2075. (in Chinese)

胡春雨,刘弘,陆佃杰.一种基于单幅图像的3D场景建模方法[J].计算机应用研究,2014,31(6):1909-1913. HU C Y, LIU H, LU D J. Monocular image based 3D-scene reconstruction method[J]. Application Research of Computers, 2014, 31(6):1909-1913. (in Chinese)

XU K, ZHENG H L, ZHANG H, et al. Photo-inspired model-driven 3D object modeling[J]. ACM Trans on Graphics, 2011, 30(4):80-89.

FELZENSZWALB PF, HUTTENLOCHER D P. Efficient graph-based image segmentation[J]. International Journal of Computer Vision, 2004, 59(2):167-181.

SAXENA A, SUN M, NG A Y. Make3D:Learning 3D scene structure from a single still image[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2008, 31(5):824-840.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于结构标签学习的显著性目标检测

基于区域特征融合的RGBD显著目标检测