Ground truth evaluation of 3D computer vision on non-rigid biological structures

M. Nielsen; H.J. Andersen; D.C. Slaughter; E. Granum

Ground truth evaluation of 3D computer vision on non-rigid biological structures

In: Precision Agriculture â05

Authors:

M. Nielsen

M. Nielsen ¹Computer Vision and Media Technology, Aalborg University, Denmark
mnielsen@cvmt.dk

Search for other papers by M. Nielsen in
Current site
Google Scholar
PubMed

H.J. Andersen

H.J. Andersen ¹Computer Vision and Media Technology, Aalborg University, Denmark

Search for other papers by H.J. Andersen in
Current site
Google Scholar
PubMed

D.C. Slaughter

D.C. Slaughter ²Biological and Agricultural Engineering, University of California Davis, USA

Search for other papers by D.C. Slaughter in
Current site
Google Scholar
PubMed

, and

E. Granum

E. Granum ¹Computer Vision and Media Technology, Aalborg University, Denmark

Search for other papers by E. Granum in
Current site
Google Scholar
PubMed

Type:: Chapter

Pages:: 547â556

DOI:: https://doi.org/10.3920/978-90-8686-549-9_068

Purchase instant access (PDF download and unlimited online access):

â¬42.00â¬35.00Â excl. VAT

Add to Cart

Redeem Access Token

There is an increasing interest in using 3D computer vision in precision agriculture. This calls for better quantitative evaluation and understanding of computer vision methods. It is difficult to evaluate the performance of an algorithm quantitatively, when dense ground truth such as dense depth map is not readily available, in contrast to sparse ground truth data at high levels of abstraction. This paper proposes a test framework using ray traced crop scenes. This enables quantitative comparison of dense 3D reconstruction algorithms, camera- and lighting setups, and post-processing before investing in expensive equipment and field experiments. In order to illustrate the possibilities in this test framework, 3240 disparity maps were computed and example hypotheses were analysed using statistics and graphs. It was to be found a valuable tool to structure the otherwise incomprehensible vast information space and to see relationships between parameter configurations and crop features.

Citation Info

Save
Cite
Email this content

Share link with colleague or librarian

You can email a link to this page to a colleague or librarian:
Email this content
or copy the link directly:

https://brill.edhh.ma/abstract/book/9789086865499/BP000068.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Collapse
Expand

Precision Agriculture â05

E-Book ISBN:: 9789086865499

Publisher:: Wageningen Academic

Print Publication Date:: 24 Oct 2005

Subjects
- Life Sciences
  - General
  - Agriculture & Environment

Metrics

	All Time	Past 365 days	Past 30 Days
Abstract Views	77	46	11
Full Text Views	0	0	0
PDF Views & Downloads	0	0	0

African Studies	Education	Media Studies
American Studies	History	Middle East and Islamic Studies
Ancient Near East and Egypt	Human Rights and Humanitarian Law	Musicology
Art History	International Law	Philosophy
Asian Studies	International Relations	Religious Studies
Biblical Studies	Jewish Studies	Slavic and Eurasian Studies
Biology	Languages and Linguistics	Social Sciences
Book History and Cartography	Life Sciences	Theology and World Christianity
Classical Studies	Literature and Cultural Studies

Subjects

Authors

Open Access

Product Information

Company

Contact

Ground truth evaluation of 3D computer vision on non-rigid biological structures

Citation Info

Share link with colleague or librarian

Precision Agriculture â05

Table of Contents

Metrics

Metrics

Precision Agriculture â05