Recently, the Japanese trade system is facing large changes due to the new Trans-Pacific Strategic Economic Partnership Agreement. To be competitive with imported goods from abroad, many industries are required to improve their production efficiency, including dairy and livestock farming. Development of new forage grass cultivars selected for high yield and quality has become an important method for increasing animal performance and productivity. Traditionally, selection of superior grass plants is based on subjective visual ratings by the breeder. Novel non-subjective selection methods have the ability to improve the accuracy and efficiency of selection. In this study, we tested whether using a UAV (drone) and image analysis can be used to enhance the efficiency and effectiveness of selection in a grass breeding program. Image analysis of a nursery containing orchard grass cultivars obtained from the UAV correlated (r > 0.8) with their first crop yield. An index based on visible wavelengths RGB showed high correlation with the degree of pathogen infection (rhynchosporium and black rust). These results indicate that combining the use of a UAV with image analysis can be helpful for the development of effective selection methods in a grass breeding program.
