Iowa State University researchers have developed a way to take 3-D pictures with a robot to study the leaf angles of corn plants.
Professor Lie Tang says the leaf angles play an important role in how the plant captures sunlight, but it’s hard for a photographer to get pictures with thousands of plants growing together. “Some of them can grow more than ten feet tall. And you want to capture the angles from bottom to the top in a very tight space,” Tang says.
Tang is a professor of agricultural and biosystems engineering and says they developed a system with multiple tiers of cameras that are customized to work together. “We have binocular vision or stereo vision, just like our human beings who have two eyes. So we can perceive the world in three-dimensional space. It’s the same kind of principle, we even have two cameras through something we call the matching between two cameras,” he says.
Tang says the cameras on the robot they call “Pheno Bot 3,” gives them color plus depth to help determine the leaf angles.
“There’s quite a breakthrough piece, to be able to exactly detect the different angles in a very complex imaging situation,” Tang says. ISU Agronomy professor Jianming Yu, is working with Tang on the project.
Tang says they use A-I or artificial intelligence and deep learning to sort through the data to measure the angles accurately.
“So we are training our computer algorithms to understand the different angles to be able to recognize different angles through training. So that has proven to be very successful,” according to Tang.
There are various corn hybrids with an upright leaf angle, to assist in the photosynthetic process and improve crop yields, but the ISU researchers say the genetics that control this trait are relatively unknown. It’s hoped the advanced photo system will help them learn more.
The research is supported by a $2.5 million-dollar National Science Foundation grant.