Face recognition AI helps save billions of dollars on the crop

GENEVA – A radical collaboration between a biologist and an engineer is working hard to protect grapes. The technology developed using robotics and AI to identify vines contaminated with fungi has recently been made available to researchers working in various botanical and animal studies.

Although biologist Lane Cadley Davidson, associate professor at the Cornell University Integrated School of Plant Sciences, is working to develop grapes that are more resistant to mildew, laboratory studies have ruled out the need to manually evaluate thousands of grapes. Samples of evidence for the disease.

The fungus, which attacks many plants, including mildew, wine, and table wine, leaves white scars on leaves and fruits, costing wine producers around the world billions of dollars a year in lost fruit and mushrooms.

Cadel Davidson is a pathogen at the US Department of Agriculture’s Agricultural Research Service. He works in the Department of Veterinary Research in Geneva, and his team developed prototype robots protocols that can automatically scan vine leaf samples: a process called phenotyping – funded by USDA-ARS through the VitisGen2 Vinegar Project and the Light and Health Research Center. This partnership led to the creation of a robot camera called the BlackBerry.

David Gaduri, a senior research fellow in the Department of Plant Pathology and Plant Microbiology at the Integrated School of Plant Sciences, uses a phonoping robot to analyze powdery mildew in hop and grape analysis.

But extracting relevant biological information from these images was still an important interest.

Enter Engineer and Computer Scientist – Yu Jiang, Assistant Research Professor in SIPS Gardening Department at Cornell Agritech. Jiang’s research focuses on system engineering, data analysis, and artificial intelligence. The BlackBerry robot can measure data at 1.2 microns per pixel – with a standard optical microscope. Each 1 cm of leaf sample is examined and the robot provides information at 8,000 by 5,000 pixels.

Extracting useful information from such a large, high-resolution image was Yanyan’s challenge, and his team used AI. Using face-to-face discoveries in computer-assisted vision systems, Jiang applied this knowledge under a microscope. Jiang and his team also applied the visualization of network transmission processes to help biologists better understand the analysis process and build trust in II models.

Working together, the Cadle-Davidson team examines and verifies what the robots see, enabling the Jiang team to teach them how to more effectively identify biological characteristics. The results are impressive, says Cadley-Davison. Six months of research experiments to complete the entire laboratory team now take BlackBerry robots in just one day.

“Our science has revolutionized,” says Cadley-Davidson. And we know that II tools really do a better job of explaining the genes of these grapes.

In July alone, the partnership received a prize and two new grants. On July 1, the team received a $ 100,000 grant from USDA-ARS to distribute to BlackBard field offices working on other high-yielding crops.

“We look forward to finding cooperative laboratories that will join us in using this tool,” Jiang said. We look at possible applications for this research in herbal studies, animal husbandry, or medical purposes.

Plant diseases such as powdery mildew can appear in infrared before they can be seen; If researchers are able to develop tools to help early detection of disease, it will allow farmers to target fungal pathogens before the spread of the infection, which includes small fungi and lost crops. They are working to integrate AI efficiently with scientists in data analysis.

“This work is accelerating the growth of wine and genetics,” said Donald Brown, president of the National Wine Research Alliance. This technology is truly shortening that time line for the benefit of farmers and consumers.


Leave a Comment