NOTE: Repeat presentations of this content will be given throughout the duration of the 10:00 - 11:30 a.m. tour, as each group rotates between tour stops.

Description of Innovation: Multi-spectral imaging utilizes statistical analysis with imaging to be able to instantaneously differentiate between seeds varying for physical and chemical traits. The system consists of a high-resolution camera that takes a series of images at wavelength bands ranging from ultraviolet to near infrared light. Because a unique image is taken at each wavelength band, the images can be viewed individually or in combination. In addition, specific layers within the image associated with a particular trait can be isolated and compared. Multivariate analysis can then be utilized to train the system to identify and sort by a particular trait in unknown samples and determine the percentage by area of each type.

Why it Matters to the Seed Industry: Multi-spectral imaging can identify differences in seeds that show know visual differences and provides an instantaneous method for sorting based on a particular trait. The system can be trained with seeds or grains of known phenotype and then be used to differentiate and sort unknown samples into phenotypic groups. Examples of traits that can be distinguished by multi-spectral imaging includes, but is not limited to, size, color, viability, disease resistance, starch content, oil content, and pre-harvest sprout. Once trained, the system could be utilized by breeding and biotech companies to make selections for new varieties, verify presence of a high value trait in developed lines, or detect violations of identity protection. In addition, the system could be implemented in the field, a grain handling facility, or a production facility to audit production efficiency and ensure the identity and purity of high value product.