BackgroundPhenotyping is a critical component of plant research. Accurate and precise trait collection, when integrated with genetic tools, can greatly accelerate the rate of genetic gain in crop improvement. However, efficient and automatic phenotyping of traits across large populations is a challenge; which is further exacerbated by the necessity of sampling multiple environments and growing replicated trials. A promising approach is to leverage current advances in imaging technology, data analytics and machine learning to enable automated and fast phenotyping and subsequent decision support. In this context, the workflow for phenotyping (image capture → data storage and curation → trait extraction → machine learning/classification → models/apps for decision support) has to be carefully designed and efficiently executed to minimize resource usage and maximize utility. We illustrate such an end-to-end phenotyping workflow for the case of plant stress severity phenotyping in soybean, with a specific focus on the rapid and automatic assessment of iron deficiency chlorosis (IDC) severity on thousands of field plots. We showcase this analytics framework by extracting IDC features from a set of ~4500 unique canopies representing a diverse germplasm base that have different levels of IDC, and subsequently training a variety of classification models to predict plant stress severity. The best classifier is then deployed as a smartphone app for rapid and real time severity rating in the field.ResultsWe investigated 10 different classification approaches, with the best classifier being a hierarchical classifier with a mean per-class accuracy of ~96%. We construct a phenotypically meaningful ‘population canopy graph’, connecting the automatically extracted canopy trait features with plant stress severity rating. We incorporated this image capture → image processing → classification workflow into a smartphone app that enables automated real-time evaluation of IDC scores using digital images of the canopy.ConclusionWe expect this high-throughput framework to help increase the rate of genetic gain by providing a robust extendable framework for other abiotic and biotic stresses. We further envision this workflow embedded onto a high throughput phenotyping ground vehicle and unmanned aerial system that will allow real-time, automated stress trait detection and quantification for plant research, breeding and stress scouting applications.Electronic supplementary materialThe online version of this article (doi:10.1186/s13007-017-0173-7) contains supplementary material, which is available to authorized users.

No abstract

Phenotypic studies require large datasets for accurate inference and prediction. Collecting plant data in a farm can be very labor intensive and costly. This paper presents the design, architecture (hardware and software) and deployment of a multi-robot system for row crop field data collection. The proposed system has been deployed in a soybean research farm at Iowa State University.Robotics 2018, 7, 61 2 of 15 Several agricultural robots have been developed in the recent past. Unlike industrial robots, the challenges for robots in agriculture are diverse. First, agricultural fields do not have a structured and controlled environment, in contrast to industrial facilities. Second, agricultural robots operate on farms with infrastructure and operating conditions different from industrial robots. Third, industrial processes can be designed by modules to complete particular tasks by a particular robot, whereas the complex tasks in agriculture sometimes cannot be split into simple actions. Due to the aforestated reasons, agricultural applications require more versatile and robust robots.One of the earliest robots deployed in agricultural applications was the German BoniRob [6]. This robot was initially developed by AMAZONEN-WERKE (Hasbergen, Germany) together with the Osnabruck University of Applied Sciences, Robert Bosch GmbH (Stuttgart, Germany) and other partners, and is now a part of Deepfield Robotics, a Bosch start-up company (Stuttgart, Germany). BoniRob was developed to eliminate some of the most tedious tasks in modern farming, plant breeding, and weeding. Another agricultural application example was RHEA [7]. It was a automatic and robotic systems with a fleet of heterogeneous ground and aerial robots developed for effective weed and pest control. Many universities have developed their own robots, for example, the multi-purpose Small Robotic Farm Vehicle (SRFV) from Queensland University of Technology, Brisbane, Australia [8]. This modular design allows the SRFV to undertake a range of agricultural tasks and experiments, including harvesting, seeding, fertilizing and weeding management. Another example is the Thorvald II [9], a completely modular platform both on the hardware and the software side. It can be reconfigured to obtain the necessary physical properties to operate in different production systems, for example, tunnels, greenhouses and open fields. Another example of a university-developed agricultural robot is the Phenobot 1.0 [10] from Iowa State University. It is an auto-steered and self-propelled field-based phenotyping platform for tall dense canopy crops.Companies that work on commercial robots exists as well. For the open field, there is the ANATIS from Carre, the Robotti from AGROINTELLI and Asterix from Adigo [11]. Ecorobotix is developing a solar powered robot for ecological and economical weeding of row crops, meadows and intercropping cultures, while Naio technologies has developed the robots OZ, BOB, TED and DINO. Other robots are designed for greenhouses. One example is the fully-auto...

No abstract

No abstract