Search
Researchers analyzed the images of rice produced from the LSU AgCenter’s breeding program to evaluate grain characteristics like chalkiness, milling and grain length at the H. Rouse Caffey Rice Research Station. Photo by Jardel da Silva Souza
LSU AgCenter scientists and genetics experts have been working on a new platform to better evaluate grain quality from rice produced through the AgCenter’s rice breeding program. The platform would allow researchers to quickly determine certain characteristics of the rice with a focus on chalk impact.
Led by Jardel da Silva Souza, a visiting researcher from Brazil, and Roberto Fritsche-Neto, an assistant professor and researcher at the H. Rouse Caffey Rice Research Station, the platform development began this past year to be able to save time and money when evaluating the rice produced. Currently, the equipment used is expensive and takes too long to adequately test the new varieties at the speed researchers would like.
Souza had done work like this before while working towards his master’s degree In Brazil. He used pictures of ornamental plants and created an algorithm to find out information on color intensity and how the different phenotypes would change a plant's architecture.
He used this idea for the new rice analysis program which uses pictures of rice to evaluate it. After the photos are put into the system, they are then processed through an algorithm and the physical characteristics of the rice, such as chalkiness, milling and grain length, can be evaluated.
The new platform has been found to be efficient at correctly evaluating the rice yields and significantly lowering the time it would take to do so. The process to look at the almost 10,000 breeding plots at the Rice Research Station would take only 10 business days and cost about $1,500, significantly less than the $40,000 currently spent on equipment.
Work is still being done to make the platform more efficient and adapt it to changing field conditions. The work will continue over the next few years to finetune the algorithm, but Souza and Fritsche-Neto are excited about the opportunity to make such a large leap forward for their research.
“As soon as I received the project, I saw enormous potential in it. It was something that needed to be done to facilitate research and speed up the process,” Souza said. “In plant breeding research, decisions need to be quick, and this project offered the opportunity to conduct applied research that could have a significant impact.”
There are still some issues with the program, like being able to find photos of the rice that can be used within the algorithm, as well as the more logistical issue of Souza returning to Brazil to finish his Ph.D. However, the program could soon be used in its intended manner increasing efficiency for researchers.
