Teilflächenspezifische Ertragserfassung bei Wurzelfrüchten (Zuckerrüben) in Echtzeit mittels Mikrowellensensoren (Radar)
Final Report Abstract
Yield mapping is a basic entity of the Precision Farming concept and provides crucial information about the success of cultivation. Several approaches to site-specific yield recording during the sugar beet harvest are known. Most of them are based on the weighing of sugar beets together with soil tare. Another real-time yield mapping approach with the option of plant population counting is based on estimating the mass of individual sugar beets on the basis of their maximal diameter. The main goal of the research was to develop and evaluate a yield recording procedure based on radar technology, which will provide non-invasive in-soil detection and identification of single sugar beets in order to enable the counting of individual sugar beets and determining of the single sugar beet root mass. Further goals were to enhance the radar technology for other applications in the agriculture, as a general goal, and to define applicability restrictions of practical utilisation of the system for the sugar beet and similar crops. The research activities have been divided into laboratory and field experiments. Laboratory experiments have been divided in the preliminary experiments and reference measurements. The preliminary experiments provided guidelines for the whole project course, especially concerning proper antenna arrangement and general conditions determination. The reference measurements with different test objects (aluminium balls, and correctly and not correctly topped sugar beets) have given valuable information about the measuring system's behaviour, which were used in field conditions and at the same time enabled the successful field measurements. The field tests have been conducted in two measuring campaigns, in August and October 2006, providing sufficient variability of test conditions, which also complied to tested laboratory conditions. Within field conditions two different scenarios have been tested: normal row with normal average distance between sugar beets of 20 cm and thinned row with increased sugar beet distance. In both cases the sugar beets have been correctly topped. The laboratory experiments showed that the larger reference test object (aluminium sphere with 0120 mm) is detectable up to the second level of soil water content of 30 ± 5 %vol, and up to the second level of the soil surface roughness determined with the standard deviation of height of the surface profile of 6 mm and maximum diameter of soil parts of 20 mm. The detectability of the smaller reference object (060 mm) was worse, and in this case the influence from the soil surface roughness was stronger. The small object was detectable up to the same level of soil water content, and up to the soil surface roughness with the standard deviation of height of the surface profile of 3 mm and maximum diameter of soil parts of 5 mm. In comparison to the influences of soil water content level and soil surface roughness level, the influence of soil texture was less significant. The laboratory test with sugar beets showed comparable results in all conditions with worse detectability because of lower dielectric contrast to the surroundings. The differentiation from the surrounding soil of sugar beets with diameter smaller than 60 mm (ca. 100 g) was not unambiguously possible in the tested laboratory conditions. On the other hand, the number of sugar beets of this size is less than 0,5 % (determined on a sample of 1770 adult sugar beets) and it represents irrelevant share of the total yield. The unambiguous differentiation of any sugar beet in the scenario with surface roughness level 3 or higher is also impossible. The used method allowed the identification and detection of 90 % to 96 % of sugar beets under test in the various field conditions, with correlation coefficients between real sugar beet positions and detected positions of about 99 %, and average positioning error from 1,1 to 3,6 cm. The correlation coefficients between single sugar beet root masses and recorded reflected energy amounts were for the majority of tests over 70%, and the best results have been on the level close to 90 %. The average sugar beet mass in August was 560 g with average diameter of 9,7 cm. The average sugar beet mass in October was more than 30 % higher. The soil water content in August was 31 %vol in average, and in October between 22,3 and 28,1 %vol. Nevertheless, it was not possible to establish a general relationship between local conditions (average sugar beet size and soil water content) and sugar beet detectability. The tests in less favourable conditions in August (smaller average sugar beet mass and higher soil water content) provided similar or in some cases even better results.
Publications
-
Konstantinovic, M., Wöckel, S., Schulze Lammers P., Sachs, J. 2006. Detektionsprinzip von Biomasse mittels UWB Radar am Beispiel von Zuckerrüben. Agrartechnische Forschung (Agricultural Engineering Research), No 12, p 93-102
-
Konstantinovic, M., Wöckel, S., Schulze Lammers P., Sachs, J. 2006. Detektionsprinzip von Biomasse mittels UWB Radar am Beispiel von Zuckerrüben. Landtechnik 61, No. 4, p 192-193
-
Konstantinovic, M., Wöckel, S., Schulze Lammers, P., Martinov, M. 2006. Elektromagnetni talasi u poljoprivredi - Detekcija biomase UWB radarom (Electromagnetic waves in the Agriculture - Biomass detection using UWB Radar). Revija agronomska saznanja, No. 3, p 8-11
-
Konstantinovic, M. 2007. In-Soil Measuring of Sugar Beet Yield Using UWB Radar Sensor System. Forschungsbericht Agrartechnik, VDI-MEG Nr. 455, pp 215
-
Konstantinovic, M., Wöckel, S., Schulze Lammers, P., Sachs, J. 2007. Influence of the sugar beet spatial arrangement on yield mapping of sugar beet using UWB radar. Book of proceedings of ECPA, Skiathos, Greece, June 3.-6
-
Konstantinovic, M., Wöckel, S., Schulze Lammers, P., Sachs, J. 2008. UWB Radar System for Yield Monitoring of Sugar Beet. Transactions of ASABE, Vol. 51 (2), pp 9
-
Konstantinovic, M., Veselinov, B., Woeckel, S., Martinov, M., Schulze Lammers, P. 2005. In-soil sugar beet yield measuring using remote sensing - early report. Contemporary agricultural engineering, Vol. 31, No. 3, p 126-135
-
Konstantinovic, M., Woeckel, S., Schulze Lammers P., J. Sachs, 2005. Yield mapping of sugar beet using ultra wideband radar - Methodology and first research results. Book of Proceedings of the VDI Conference Agricultural Engineering, Hanover, Germany, p 497-502
-
Konstantinovic, M., Wöckel, S., Schulze Lammers P., Sachs, J. 2006. Concept and features of an ultra wideband radar system for mapping sugar beet yield. CIGR-EurAGEng-VDI-MEG Book of Abstracts of the World Congress, September 3-7, Bonn, Germany, p 353-354
-
Konstantinovic, M., Wöckel, S., Schulze Lammers, P., Sachs, J. 2007. Evaluation of a UWB Radar System for Yield Mapping of Sugar Beet. ASABE Paper Number: 071051. St. Joseph, Mich. USA, pp
-
Konstantinovic, M., Wöckel, S., Schulze Lammers, P., Sachs, J., and Martinov, M. 2007. Detection of Root Biomass using Ultra Wideband Radar - an Approach to Potato Nest Positioning. Agricultural Engineering International: the CIGR Ejournal. Manuscript IT 06 003. Vol. IX. May, 2007
-
M. Konstantinovic, Project presentation at the China Agricultural University (CAU) Centre of Precision Agriculture, Beijing, China with the Title: Radar Application in the Agriculture with the example of the sugar beet detection system, Oktober 2005
-
Wöckel, S., Konstantinovic, M., Sachs, J., Schulze Lammers, P., Kmec, M. 2006. Application of ultra-wideband M-Sequence-Radar to detect sugar beets in agricultural soils. 11th international Conference on Ground Penetrating Radar, June 19-22, Columbus Ohio, USA.
