Search Site   
Current News Stories
Solar eclipse, new moon coming April 8
Mystery illness affecting dairy cattle in Texas Panhandle
Teach others to live sustainably
Gun safety begins early
Hard-cooked eggs recipes great for Easter, anytime
Michigan carrot producers to vote on program continuation
Suggestions to celebrate 50th wedding anniversary
USDA finalizes new ‘Product of the USA’ labeling rule 
U.S. weather outlooks currently favoring early planting season
Weaver Popcorn Hybrids expanding and moving to new facility
Role of women in agriculture changing Hoosier dairy farmer says
   
News Articles
Search News  
   

Iowa researchers develop app for soybean diseases

 

By DOUG SCHMITZ

Iowa Correspondent

AMES, Iowa — An independent team of Iowa State University agronomists and engineers have developed a mobile device app that could allow farmers – or anyone – with access to a smart phone detect soybean diseases in a similar way trained plant breeders and scientists already do.

“We want this technology to allow machines to see with the eyes of an experienced plant breeder,” said Arti Singh, an ISU adjunct assistant professor of agronomy, who led researchers to start collecting a large dataset of approximately 25,000 images of Iowa soybean stresses.

Singh said the result is a computer application that can diagnose and quantify the amount of various foliar stresses by analyzing digital images of soybeans, improving efficiency for plant breeders and farmers, while demonstrating the growing value of automation in agriculture.

Published in the April 16 edition of the widely peer-reviewed journal Proceedings of the National Academy of Sciences of the United States of America, she said the findings demonstrate how artificial intelligence can identify a range of common stresses in soybeans.

“Plant stress identification based on visual symptoms has predominately remained a manual exercise performed by trained pathologists, primarily due to the occurrence of confounding symptoms,” the researchers wrote in their article. “However, the manual rating process is tedious, is time-consuming and suffers from inter- and intra-rater (degree of agreement among multiple tests performed by a rater) variabilities.”

The team said their work resolves such issues via the concept of explainable deep machine learning to automate the process of plant stress identification, classification and quantification.

“We construct a very accurate model that can not only deliver trained pathologist-level performance but can also explain which visual symptoms are used to make predictions,” the researchers wrote. “We demonstrate that our method is applicable to a large variety of biotic and abiotic stresses and is transferable to other imaging conditions and plants.”

The research was supported by funding from the Iowa Soybean Assoc., the ISU Plant Sciences Institute, the ISU Presidential Initiative for Interdisciplinary Research, the USDA and the National Science Foundation. As part of the research, the team developed an automated machine-learning framework to find patterns in the soybean leaf images that correlated with eight common sources of stress, such as diseases, nutrient deficiency and herbicide injury.

According to Singh, scouting crops and conducting visual measurements for stress is a time-consuming and often inconsistent process both for plant breeders and farmers. Introducing an automated tool can save time and produce more standardized results.

Soumik Sarkar, an ISU assistant professor of mechanical engineering and co-author, said the researchers compared the performance of their program with actual diagnoses from trained plant scientists, and the results showed excellent correlation. In addition, Sarkar said the program qualifies its diagnosis by providing the specific visual symptoms it noted to reach its conclusions.

While the program is currently available only for academic communities, the researchers said they intend to deploy the application on smart phones to make the product available widely. The technology also has the potential for use in unmanned aerial vehicles and ground robots.

“This is a prime example of how artificial intelligence can be applied to agriculture,” Sarkar said. “It can provide more automation and more efficiency than the traditional way of diagnosing these stresses.”

In August 2016, the University of Tennessee Institute of Agriculture’s Department of Entomology and Plant Pathology released a mobile-friendly disease field guide – the UT Field Crops Guide. It’s designed to help growers, agents and consultants quickly assess foliar diseases in their soybean fields.

Heather Young-Kelly, UT assistant professor of plant pathology and Southern Integrated Pest Management Center coordinator, who designed the new guide, said her goal – similar to ISU’s mission – was to create “an informative site that would look great and function well on the devices most growers are using – smart phones.

“I believe more and more growers are using the UT guide to help diagnose diseases and insect pests, as well as provide management options,” she explained. “Since its launch, the site has had 15,776 page views and 11,283 unique page views (which essentially means 4,500 repeat users).

“The field guide was designed to take the hard-copy print guides we had and make them available in a mobile-friendly version, so when in the field, the essentials could be pulled up on grower’s phone, identification information, threshold and management options, including insecticide and fungicide efficacy tables. But more technically, the site was designed in WordPress in 2016, and further developed and updated since.”

In addition, the guide features general information on all foliar diseases that pose an economic threat to Tennessee’s soybean crop. According to Young-Kelly, proper identification is made easier through video instruction and a comprehensive photo gallery.

She said these features allow users to easily compare what they are seeing on the screen with what they are seeing in the field. Once the problem is identified, users can view management options for each disease, including information on variety selection, fungicide efficacy and resistance.

“After the original kickoff of diseases of soybeans, the insect cotton portion was added, as well as additional disease sections for wheat and cotton,” she said. “I believe in 2018, foliar diseases of corn and soybean insects will be added to the site and continuous update of the site will continue as needed. I’ve gotten very positive feedback on the field guide.”

She said the UT field guide online at http://guide.utcrops.com is very different from ISU’s new app. “With it, you take a picture and it is supposed to identify the stresses,” she noted. “It seems, from my perspective, this technology is still a ways out from being available to farmers for use.”

In fact, while Kelly has been able to use imagery that correlates to her disease ratings, she said the imagery was not able to discern between different diseases or injuries. “In my opinion, one still needs to ground-truth the stress in the field, which is where the UT guide could help after imagery identifies an area of concern showing stress.”

The ISU interdisciplinary research team also included Asheesh Singh, an associate professor of agronomy, and Baskar Ganapathysubramanian, an associate professor of mechanical engineering. Two ISU Ph.D. students – Sambuddha Ghosal in mechanical engineering and David Blystone in agronomy – also contributed.

7/18/2018
Iowa researchers develop app for soybean diseases

' st_image='http://www.farmworldonline.com/Newsimages/F_logo.png'> Iowa researchers develop app for soybean diseases

' st_image='http://www.farmworldonline.com/Newsimages/F_logo.png'> Iowa researchers develop app for soybean diseases

' st_image='http://www.farmworldonline.com/Newsimages/F_logo.png'> Iowa researchers develop app for soybean diseases

' st_image='http://www.farmworldonline.com/Newsimages/F_logo.png'> Iowa researchers develop app for soybean diseases

' st_image='http://www.farmworldonline.com/Newsimages/F_logo.png'>