Abstract
Witchweeds (Striga spp.) and broomrapes (Orobanchaceae and Phelipanche spp.) are root parasitic plants that infest many crops in warm and temperate zones, causing enormous yield losses and endangering global food security. Seeds of these obligate parasites require rhizospheric, host-released stimulants to germinate, which opens up possibilities for controlling them by applying specific germination inhibitors or synthetic stimulants that induce lethal germination in the host's absence. To determine their effect on germination, root exudates or synthetic stimulants/inhibitors are usually applied to parasitic seeds in in vitro bioassays, followed by assessment of germination ratios. Although these protocols are very sensitive, the germination recording process is laborious, representing a challenge for researchers and impeding high-throughput screens. Here, we developed an automatic seed census tool to count and discriminate germinated seeds (GS) from non-GS. We combined deep learning, a powerful data-driven framework that can accelerate the procedure and increase its accuracy, for object detection with computer vision latest development based on the Faster Region-based Convolutional Neural Network algorithm. Our method showed an accuracy of 94% in counting seeds of Striga hermonthica and reduced the required time from approximately 5 min to 5 s per image. Our proposed software, SeedQuant, will be of great help for seed germination bioassays and enable high-throughput screening for germination stimulants/inhibitors. SeedQuant is an open-source software that can be further trained to count different types of seeds for research purposes.
Highlights
Root parasitic weeds, such as witchweeds (Striga spp.) and broomrapes (Orobanchaceae and Phelipanche spp.), are one of the major biological threats to the production of major agricultural food crops (Tank et al, 2006; Musselman et al, 2001; Parker, 2012; Pennisi, 2010; Rodenburg et al, 2016), as infestation by these obligate parasites causes yield losses ranging from a few percent to complete crop failure (Gressel et al, 2004; Ejeta, 2007; Atera et al, 2012)
Faster R-convolution neural network (CNN) is an efficient two-stage method consisting of a Region Proposal Network (RPN) and an object detector (R-CNN) that takes raw images as input and extracts a meaningful feature representation based on a Residual Network backbone (ResNet)
The ResNeXt-101-Feature Pyramid Network (FPN) (Xie et al, 2017) is an enhanced version of ResNet, which performs a set of transformations inside residual blocks, allowing the network to increase its accuracy while preserving the same number of parameters
Summary
Root parasitic weeds, such as witchweeds (Striga spp.) and broomrapes (Orobanchaceae and Phelipanche spp.), are one of the major biological threats to the production of major agricultural food crops (Tank et al, 2006; Musselman et al, 2001; Parker, 2012; Pennisi, 2010; Rodenburg et al, 2016), as infestation by these obligate parasites causes yield losses ranging from a few percent to complete crop failure (Gressel et al, 2004; Ejeta, 2007; Atera et al, 2012) They jeopardize global agriculture due to their variety of hosts (Xie et al, 2010): witchweeds attack cereal crops in subSaharan Africa (Gressel et al, 2004; Parker, 2012), while broomrapes infest non-cereal crops in Central Asia and the Mediterranean area (Joel et al, 2007; Parker, 2012). This allows the parasites to grow, break the soil surface, and continue their above-ground development to reach maturity: a single parasitic plant can produce tens of thousands of tiny and highly viable seeds that return into the soil and supply an already huge seedbank in constant expansion (Ejeta, 2007; Jamil et al, 2012)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
