2019/06 – 2021/02, EFRE-funded Start-Up, TRANSFER.NRW
Accelerating and improving breeding towards more efficient crops and varieties is key for increasing yield and improving the resilience of plants. For plant breeders, it is important to observe and document phenotypic traits describing the appearance of plants in the field to evaluate the quality and success of the breeding process. With Pheno-Inspect, we aim at offering growers and farmers novel software solutions for automated high-throughput phenotyping in the fields. For remote sensing, we rely on small and lightweight aerial platforms, which enable a flexible, large-area, and time-efficient survey of fields and plot experiments. With our software toolchain, we provide breeders and farmers with a tool to gain precise knowledge of the crop and individual plants. We automatically detect phenotypic traits of crop plants, the species of plants, and weeds in the field and derive site- or plot-specific statistics. Our approach relies on state-of-the-art machine learning methods optimized for the agricultural domain to interpret the captured image data semantically and to extract the desired parameters about the plants. The learning procedures developed by us use expert knowledge inserted by the user to adapt efficiently and thus deliver the desired results quickly and effectively with regard to individual problems and local characteristics of the environment.
This project is funded by the European Union and the State of North Rhine-Westphalia
2021/03 – 2024/02, BLE & BMEL
RegisTer targets the use of artificial intelligence and optical sensors in the variety description of the Federal Plant Variety Office in the examination for distinctiveness, uniformity, and stability and the value for cultivation within the scope of the variety approval for sugar beets.
Sugar beet is a common crop in Germany and represents an important economic factor for rural areas. Modern varieties must have diverse characteristics such as disease, stress tolerance, and high yield potential. The breeding and approval process must recognize these characteristics quickly and reliably. Furthermore, each variety must be clearly described and distinguishable. The interdisciplinary collaborative project RegisTer aims to develop automated routines for the characterization and evaluation of sugar beet varieties based on the plants’ geometric and optical/reflective properties, which are measured using ultra-light flying drones. Drones, equipped with high-resolution RGB, multi-spectral cameras, and 3D sensors, measure test fields. A millimeter-precise data set is the basis for automatically analyzing plant parameters with state-of-the-art image processing based on machine learning down to the individual plant level. The aim is to automatically extract (new) plant characteristics and their valuable properties for variety description and evaluation. This project aims to achieve automation, standardization, and improvement of the assessment process for the register and value tests at the Federal Plant Variety Office and performance tests in the plant breeding industry, which small and medium-sized enterprises characterize.
The RegisTer project is technologically based on recording experimental plots and individual plants using drones and terrestrial laser scanning. The drone, equipped with high-resolution RGB and multi-spectral cameras, systematically records test plots at various locations within Germany and observes the plants with the required resolution in the millimeter range. These data form the basis for the investigation of the phenotype. In addition to the 3D data from drone images’ processing, we also collect high-precision 3D data utilizing terrestrial laser scanning, which we use to extract the leaf apparatus’ geometric parameters and for quality control of the 3D data from the drone images.
For the extraction of the register and value characteristics, we develop the most modern image processing with machine learning methods, which aim to recognize and analyze single plants, but at the same time also for the evaluation on the level of the plots. In addition to automatically recognizing already existing traits, we also examine the data of new, previously unused factors for register inspection. For this purpose, we pursue the working hypothesis that the data contain variety-specific patterns suitable for the distinctness and identification of sugar beet varieties. We link the data over different points in time and derive dynamic characteristics. We investigate these parameters to develop new traits for variety description and use the temporal correlation in the data to improve the self-learning algorithms further.
RegisTer is partially supported by funds of the Federal Ministry of Food and Agriculture (BMEL) based on a decision of the Parliament of the Federal Republic of Germany. The Bundesanstalt für Landwirtschaft und Ernährung / Federal Office of Agriculture and Food in Germany (BLE) provides coordinating support for digitalization in agriculture as funding organization, grant number FZK 28DK108A20.
Herbicide-Free Weed Control
2021/09 – 2023/12, DBU
Where weeds grow, we spray. Chemical weed control is still common on about 90% of all agriculturally used land in Germany. Numerous studies prove the connection between pesticide use and the significant decline in biodiversity.
In comparison, organic farming leads to a significant increase in biodiversity and thus makes an essential contribution to a functioning ecosystem. Non-chemical weed control is often more laborious and expensive, as manual weeding is still indispensable here, significantly when growing sown, slow-growing crops such as carrots and other root crops.
The start-up tiefgrün precision weeding aims to bring an ecological method to practical maturity based on self-learning plant recognition, which can be economically competitive with herbicide application in the long term. Pheno-Inspect GmbH is responsible for developing the necessary image-based classification of plants and weeds.
A project goal is a demonstration machine for automated weed control in organic carrot cultivation: the KelvinR370. The machine is designed to detect plants based on cameras, distinguish crops from weeds with a high degree of certainty, and precisely regulate the latter. The project comprises the development, construction, and field testing of a prototype and, building on this, the implementation and practical evaluation of the KelvinR370 demonstration machine by tiefgrün. For use in sensitive crop stands, tiefgrün is developing a precise, pinpoint weed control system, integrating it into the device, and investigating its effect under practical conditions. At the same time, Pheno-Inspect is developing an image processing pipeline tailored to the system for robust classification of carrots and weeds and simultaneous detection of their root points – with maximum reliability under different field conditions.
Sponsored by the Deutsche Bundesstiftung Umwelt
TOMEYE – The first fully automated aerial crop monitoring solution for greenhouse horticulture
Crop monitoring is essential to give growers insight into crop health and development. Yield forecasts for fresh produce are fundamental to optimizing daily operations and supply chain management. Currently, monitoring is mainly done by hand, which is labor-intensive, costly, and inaccurate. TOMEYE will significantly improve yield forecasting without manual work, considerably reducing farmer income loss and labor costs. The consortium will introduce TOMEYE, a revolution in the autonomous monitoring indoor crops. TOMEYE combines novel agile indoor drone navigation with artificial intelligence for yield prediction in tomatoes, the EU’s largest market for greenhouse crops. Expansion to other crops and services related to plant development and diseases will follow.