A biotechnology company, now a division of Bayer, needed to image plants to analyze the phenotypic effects of thousands of genetic modification experiments.  The images needed to be uniform between stations and over a period of years. BME Systems designed and fabricated the semi-automated stations that included calibrated color and lighting.

Dozens of features of the plants were required for each of hundreds of plants imaged each day.  BME Systems developed a machine vision system to analyze the rosette of the plant automatically to objectively extract dozens of metric parameters such as number of stems, leaf size and shape, stem and leaf distribution and biomass from difficult to assess features such as brown leaflets against a soil background.

Note:  AI wasn’t mature at the time this project was created, but this is a great example of a system that can be built to generate data for AI training.