Plant tracking-based motion analysis in a crop field
Some agricultural tasks performed in a crop field consist of applying chemical treatment to the plants. In order to automate these tasks, a vision system and a treating device can be used. A camera looks at the plants, and the vision system tracks each plant in a sequence of images, so that it can be known the position of each one with respect to the treating device. The exact positions of each plant in the field have to be found, and the position of the vehicle, that is, a map of the field has to be recovered. This implies to track the plants, to recover the motion parameters of the vehicle and to place the plants in the field. A method to identify the plants and a shape description to match them are presented. A Hough Transform-like technique is used to select matches that come from the same movement. Motion parameters are recovered from plant correspondences using a method to find the motion of a planar patch. A Kalman filter is used to integrate the different observations of each plant. Results using real images are presented including zigzag and rotational movement.
Unable to display preview. Download preview PDF.
- M. Turk, K. D. Morgenthaler, K. D. Gremban, M. Marra,“A vision system for autonomous land vehicle navigation”, IEEE Trans. P.A.M.I. (10), pp. 342–360, 1988.Google Scholar
- S. Tsuji, J. Y. Zheng, “Visual path planning”, Proceedings of the Xth. International Joint Conference on Artificial Intelligence (2), pp. 1127–1130, 1987.Google Scholar
- L. Matthies, T. Kanade, R. Szeliski, “Kalman filter-based algorithms for estimating depth from image sequences”, International Journal on Computer Vision (3), pp. 209–236, 1989.Google Scholar
- J. Boberg, T. Salakosi, “General formulation and evaluation of agglomerative clustering methods with metric and non-metric distances”, Pattern Recognition (26),pp 1395–1406,1993.Google Scholar
- P. Trahanias, E. Skordalaskis, “An efficient sequential clustering method”, Pattern Recognition (22), pp. 449–453, 1989.Google Scholar
- J. G. Leu, “Computing a shape's moments from its boundary”, Pattern Recognition (24), pp. 949–957, 1991.Google Scholar
- O. Faugueras, “Three-Dimensional Computer Vision, a Geometric Viewpoint”, Cambridge (MA), 1993.Google Scholar
- R. Y. Tsai, T. S. Huang, “Estimating 3-D motion parameters of a rigid planar patch”, IEEE Trans. on Acoustics, Speech and Signal Processing, (29), pp. 1147–1152, 1981.Google Scholar