Development of autonomous robotic platforms for sugar beet crop phenotyping using artificial vision

Phenotyping is a major challenge in international agronomic competition. In a perspective of modern and sustainable agriculture, understanding the relationship between genotype and phenotype according to the environment is one of the major projects of agronomic research. Artificial vision devices embedded on robotic platforms, working in visible or hyperspectral color fields permit to carry out many geometric and colorimetric measurements on crops. From this information, operations like crop varieties comparisons and disease detection are realized. For sugar beet crops, phenotyping operations are made from a two small leaf stage up to the final stage just before harvesting task. Two robotic devices were used to make colorimetric and geometrical measurements on sugar beet plants. An autonomous mobile robot navigating in crop lines for little growth stages, embedded two cameras. A first one, with an oblique orientation permitted to realize autonomous crop raw tracking and the second one in a vertical position was used to record cartographic images and make detailed measurements on sugar beet plants. The second robotic platform was a manipulator arm with 6 degrees of freedom, fixed on a mobile linear axis to make measurements for advanced growing stages. Active perception operations realized with the embedded camera fixed at its extremity, consisted in locating by artificial vision the plant leaves in 3D environment and from this information, the camera was automatically positioned at various desired heights and orientations for each detected leaf, for carrying out, with accuracy, image acquisitions and measurements. Experimentations realized with both robotic platforms, for various sugar beet growing stages, shown the interest of these devices for following and analyzing in detail the geometric and colorimetric evolution of sugar beet plants in the fields, in order to carry out some phenotyping measurements and particularly for detecting some diseases.

[1]  Bernard Bayle,et al.  Manipulability of Wheeled Mobile Manipulators: Application to Motion Generation , 2003 .

[2]  R. Klose,et al.  Usability study of 3 D Time-of-Flight cameras for automatic plant phenotyping , 2009 .

[3]  Hanno Scharr,et al.  Leaf segmentation in plant phenotyping: a collation study , 2016, Machine Vision and Applications.

[4]  Gonzalo Pajares,et al.  Automatic detection of crop rows in maize fields with high weeds pressure , 2012, Expert Syst. Appl..

[5]  F. Truchetet,et al.  Crop/weed discrimination in perspective agronomic images , 2008 .

[6]  丁伟利 Ding Weili,et al.  Vanishing Point Detection Algorithm for Urban Road Image Based on the Envelope of Perpendicular and Parallel Lines , 2014 .

[7]  Tsuneo Yoshikawa,et al.  Analysis and Control of Robot Manipulators with Redundancy , 1983 .

[8]  John F. Reid,et al.  Hough-transform-based vision algorithm for crop row detection of an automated agricultural vehicle , 2005 .

[9]  Hui Kong,et al.  Generalizing Laplacian of Gaussian Filters for Vanishing-Point Detection , 2013, IEEE Transactions on Intelligent Transportation Systems.

[10]  Yunong Zhang,et al.  QP-based refined manipulability-maximizing scheme for coordinated motion planning and control of physically constrained wheeled mobile redundant manipulators , 2016, Nonlinear Dynamics.

[11]  Antonio M. Pascoal,et al.  Adaptive, non-singular path-following control of dynamic wheeled robots , 2003, 42nd IEEE International Conference on Decision and Control (IEEE Cat. No.03CH37475).

[12]  Jayme Garcia Arnal Barbedo,et al.  Digital image processing techniques for detecting, quantifying and classifying plant diseases , 2013, SpringerPlus.

[13]  Albert-Jan Baerveldt,et al.  A mobile robot for mechanical weed control , 2003 .

[14]  Anne-Katrin Mahlein,et al.  Hyperspectral imaging for small-scale analysis of symptoms caused by different sugar beet diseases , 2012, Plant Methods.

[15]  Clive H. Bock,et al.  Plant Disease Severity Estimated Visually, by Digital Photography and Image Analysis, and by Hyperspectral Imaging , 2010 .

[16]  D. Reddy,et al.  Use of Direct Antigen Coating and Protein A Coating ELISA Procedures for Detection of Three Peanut Viruses. , 1987 .

[17]  Nicola J. Ferrier,et al.  A high throughput robot system for machine vision based plant phenotype studies , 2012, Machine Vision and Applications.

[18]  J M Henson,et al.  The polymerase chain reaction and plant disease diagnosis. , 1993, Annual review of phytopathology.

[19]  Reza Ehsani,et al.  Review: A review of advanced techniques for detecting plant diseases , 2010 .