1A. Ruckelshausen, 2K. V. Alheit, 1L. Busemeyer, 1R. Klose, 1A. Linz, 1K. Moeller, 3F. Rahe, 1M. Thiel, 1D. Trautz, 4U. Weiss
1. University of Applied Sciences Osnabr
2. University of Hohenheim, Landessaatzuchtanstalt; Fruwirthstra
3. Amazonen-Werke H. Dreyer GmbH & Co. KG ; Postfach 51 ; 49202 Hasbergen-Gaste; Germany
4. Robert Bosch GmbH; Corporate Sector Research and Advance Engineering (CR-AE32); Postfach 30 02 40 ; 70442 Stuttgart; Germany
Sensor and system technologies are key components for automatic treatment of individual plants as well as for plant phenotyping in field trials. Based on experiences in research and application of sensors in agriculture the authors have developed phenotyping platforms for field applications including sensors, system and software development and application-specific mountings.
Sensor and data fusion have a high potential by compensating varying selectivities of single sensors. As for example, using the third dimension with 3D-Time-of-flight cameras has a potential for higher selectivities as compared to 2D-cameras. Spectral imaging allows a pointwise spectral analysis and is thus superior to standard NIR spectrometers. Furthermore light-curtain imaging, color cameras and optical distance sensors are used for measuring morphological plant parameters. For online detection and data storage a modular and flexible system architecture has been designed and realized including Gigabit Ethernet, a data base, error handling and human-machine interfaces.
Two very different application-driven setups of phenotyping platforms have been developed: For triticale (×Triticosecale Wittmack) a trailer allows measurement of high-density plants for growth stages up to 1.50 m. Properties of single plants and plot-related values are detected for offline analysis. For corn (Zea mays) the autonomous crop scout “BoniRob” has been developed. The robot is equipped with a RTK-DGPS allowing the re-detection of a single plant in a later run. Thus the growing process of single plants can be evaluated, a revolution for field trials. For robot navigation additional online sensor and data interpretation is applied. The sensor and system technologies are evaluated in laboratory carousels and band-conveyor.
For both platforms first test measurements have been performed, first field trials will be performed in early 2010. The work is supported by the German Federal Ministry of Food, Agriculture and Consumer Protection (project “BoniRob”) and the Federal Ministry of Education and Research (project “BreedVision”).
