Browsing by Subject "Plant phenotyping"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Publication
    Crop plant reconstruction and feature extraction based on 3-D vision
    (2019) Vázquez Arellano, Manuel; Griepentrog, Hans
    3-D imaging is increasingly affordable and offers new possibilities for a more efficient agricul-tural practice with the use of highly advances technological devices. Some reasons contrib-uting to this possibility include the continuous increase in computer processing power, the de-crease in cost and size of electronics, the increase in solid state illumination efficiency and the need for greater knowledge and care of the individual crops. The implementation of 3-D im-aging systems in agriculture is impeded by the economic justification of using expensive de-vices for producing relative low-cost seasonal products. However, this may no longer be true since low-cost 3-D sensors, such as the one used in this work, with advance technical capabili-ties are already available. The aim of this cumulative dissertation was to develop new methodologies to reconstruct the 3-D shape of agricultural environment in order to recognized and quantitatively describe struc-tures, in this case: maize plants, for agricultural applications such as plant breeding and preci-sion farming. To fulfil this aim a comprehensive review of the 3-D imaging systems in agricul-tural applications was done to select a sensor that was affordable and has not been fully inves-tigated in agricultural environments. A low-cost TOF sensor was selected to obtain 3-D data of maize plants and a new adaptive methodology was proposed for point cloud rigid registra-tion and stitching. The resulting maize 3-D point clouds were highly dense and generated in a cost-effective manner. The validation of the methodology showed that the plants were recon-structed with high accuracies and the qualitative analysis showed the visual variability of the plants depending on the 3-D perspective view. The generated point cloud was used to obtain information about the plant parameters (stem position and plant height) in order to quantita-tively describe the plant. The resulting plant stem positions were estimated with an average mean error and standard deviation of 27 mm and 14 mm, respectively. Additionally, meaning-ful information about the plant height profile was also provided, with an average overall mean error of 8.7 mm. Since the maize plants considered in this research were highly heterogeneous in height, some of them had folded leaves and were planted with standard deviations that emulate the real performance of a seeder; it can be said that the experimental maize setup was a difficult scenario. Therefore, a better performance, for both, plant stem position and height estimation could be expected for a maize field in better conditions. Finally, having a 3-D re-construction of the maize plants using a cost-effective sensor, mounted on a small electric-motor-driven robotic platform, means that the cost (either economic, energetic or time) of gen-erating every point in the point cloud is greatly reduced compared with previous researches.
  • Thumbnail Image
    Publication
    Development and assessment of a multi-sensor platform for precision phenotyping of small grain cereals under field conditions
    (2014) Busemeyer, Lucas; Würschum, Tobias
    The growing world population, changing food habits especially to increased meat consumption in newly industrialized countries, the growing demand for energy and the climate change pose major challenges for tomorrows agriculture. The agricultural output has to be increased by 70% by 2050 to achieve food and energy security for the future and 90% of this increase must be achieved by increasing yields on existing agricultural land. Achieving this increase in yield is one of the biggest challenges for the global agriculture and requires, among other things, an efficient breeding of new, higher-yielding varieties adapted to the predicted climate change. To achieve this goal, new methods need to be established in plant breeding which include efficient genotyping and phenotyping approaches of crops. Enormous progress has been achieved in the field of genotyping which enables to gain a better understanding of the molecular basis of complex traits. However, phenotyping must be considered as equally important as genomic approaches rely on high quality phenotypic data and as efficient phenotyping enables the identification of superior lines in breeding programs. In contrast to the rapid development of genotyping approaches, phenotyping methods in plant breeding have changed only little in recent decades which is also referred to as phenotyping bottleneck. Due to this discrepancy between available phenotypic and genotypic information a significant potential for crop improvement remains unexploited. The aim of this work was the development and evaluation of a precision phenotyping platform for the non-invasive measurement of crops under field conditions. The developed platform is assembled of a tractor with 80 cm ground clearance, a carrier trailer and a sensor module attached to the carrier trailer. The innovative sensors for plant phenotyping, consisting of several 3D Time-of-Flight cameras, laser distance sensors, light curtains and a spectral imaging camera in the near infrared reflectance (NIR) range, and the entire system technology for data acquisition were fully integrated into the sensor module. To operate the system, software with a graphical user interface has been developed that enables recording of sensor raw data with time- and location information which is the basis of a subsequent sensor and data fusion for trait determination. Data analysis software with a graphical user interface was developed under Matlab. This software applies all created sensor models and algorithms on sensor raw data for parameter extraction, enables the flexible integration of new algorithms into the data analysis pipeline, offers the opportunity to generate and calibrate new sensor fusion models and allows for trait determination. The developed platform facilitates the simultaneous measurement of several plant parameters with a throughput of over 2,000 plots per day. Based on data of the years 2011 and 2012, extensive calibrations were developed for the traits plant height, dry matter content and biomass yield employing triticale as a model species. For this purpose, 600 plots were grown each year and recorded twice with the platform followed by subsequent phenotyping with state-of-the-art methods for reference value generation. The experiments of each year were subdivided into three measurements at different time points to incorporate information of three different developmental stages of the plants into the calibrations. To validate the raw data quality and robustness of the data collection and reduction process, the technical repeatability for all developed data analysis algorithms was determined. In addition to these analyses, the accuracy of the generated calibrations was assessed as the correlations between determined and observed phenotypic values. The calibration of plant height based on light curtain data achieved a technical repeatability of 0.99 and a correlation coefficient of 0.97, the calibration of dry matter content based on spectral imaging data a of 0.98 and a of 0.97. The generation and analysis of dry biomass calibrations revealed that a significant improvement of measurement accuracy can be achieved by a fusion of different sensors and data evaluations. The calibration of dry biomass based on data of the light curtains, laser distance sensors, 3D Time-of-Flight cameras and spectral imaging achieved a of 0.99 and a of 0.92. The achieved excellent results illustrate the suitability of the developed platform, the integrated sensors and the data analysis software to non-invasively measure small grain cereals under field conditions. The high utility of the platform for plant breeding as well as for genomic studies was illustrated by the measurement of a large population with a total of 647 doubled haploid triticale lines derived from four families that were grown in four environments. The phenotypic data was determined based on platform measurements and showed a very high heritability for dry biomass yield. The combination of these phenotypic data with a genomic approach enabled the identification of quantitative trait loci (QTL), i.e., chromosomal regions affecting this trait. Furthermore, the repeated measurements revealed that the accumulation of biomass is controlled by temporal genetic regulation. Taken together, the very high robustness of the system, the excellent calibration results and the high heritability of the phenotypic data determined based on platform measurements demonstrate the utility of the precision phenotyping platform for plant breeding and its enormous potential to widen the phenotyping bottleneck.