African Journal of
Agricultural Research

  • Abbreviation: Afr. J. Agric. Res.
  • Language: English
  • ISSN: 1991-637X
  • DOI: 10.5897/AJAR
  • Start Year: 2006
  • Published Articles: 6693

Full Length Research Paper

Modeling of rapeseed at maturity stage using 3D unorganized point clouds and digital images

Ruifang ZHAI
  • Ruifang ZHAI
  • Research Institute for Computer Applications, Huazhong Agricultural University, Wuhan, 430070, P. R. China.
  • Google Scholar
Xiu JING
  • Xiu JING
  • Research Institute for Computer Applications, Huazhong Agricultural University, Wuhan, 430070, P. R. China.
  • Google Scholar
Chengda LIN
  • Chengda LIN
  • Research Institute for Computer Applications, Huazhong Agricultural University, Wuhan, 430070, P. R. China.
  • Google Scholar
Hui PENG
  • Hui PENG
  • Research Institute for Computer Applications, Huazhong Agricultural University, Wuhan, 430070, P. R. China.
  • Google Scholar
Jun LUO
  • Jun LUO
  • Research Institute for Computer Applications, Huazhong Agricultural University, Wuhan, 430070, P. R. China.
  • Google Scholar


  •  Received: 12 April 2011
  •  Accepted: 30 April 2014
  •  Published: 17 June 2014

Abstract

Creating 3D plant models is often a difficult and laborious task. To make it easier and more natural, the integration of digital images and 3D unorganized point clouds from a digitizer provides a promising approach for rapeseed model generation. In the present study, 3D unorganized point clouds and digital images were incorporated in the generation of complex models of rapeseeds at maturity stage. Unorganized point clouds and image sequences were taken from different viewpoints using a 3D digitizer. The 3D unorganized points and image sequences were used for the automated registration of all data sets from all the viewpoints, which is pair-wise registration. Later, all the pair-wise registration parameters were used as initial transformation parameters for multiple registrations. The next procedure generated a surface model by triangulated irregular network using all the point clouds. The capabilities of our system were demonstrated through real data sets. Experimental results showed that the average normal distances between the two scans were less than 0.3 mm after simultaneous registration, which indicated that the proposed methodology is effective and efficient.

 

Key words: Rapeseed, 3D modeling, unorganized point clouds.