Classification of Plant Structures from Uncalibrated Image Sequences
Abstract
This paper demonstrates the feasibility of recovering fine-scale plant structure in 3D point clouds by leveraging recent advances in structure from motion and 3D point cloud segmentation techniques. The proposed pipeline is designed to be applicable to a broad variety of agricultural crops. A particular agricultural application is described, motivated by the need to estimate crop yield during the growing season. The structure of grapevines is classified into leaves, branches, and fruit using a combination of shape and color features, smoothed using a conditional random field (CRF). Our experiments show a classification accuracy (AUC) of 0.98 for grapes prior to ripening (while still green) and 0.96 for grapes during ripening (changing color), significantly improving over the baseline performance achieved using established methods
This work was done by Debadeepta Dey while he was a Research Intern at Intel Research, Pittsburgh in the summer of 2010.
BibTeX
@workshop{Dey-2012-7425,author = {Debadeepta Dey and Lily Mummert and Rahul Sukthankar},
title = {Classification of Plant Structures from Uncalibrated Image Sequences},
booktitle = {Proceedings of IEEE Workshop on the Applications of Computer Vision (WACV '12)},
year = {2012},
month = {January},
editor = {Anderson Rocha},
pages = {329 - 336},
publisher = {IEEE},
keywords = {Structure from Motion, robotics, agriculture, 3D reconstruction},
}