디지털육종을 위한 RGB 이미지 기반 사과 과실 형태 측정 최적화 연구

Optimization Study of RGB Image-based Apple Fruit Measurement for Digital Breeding

The use of digital cameras in plant phenotyping studies using RGB sensors has increased. However, the need for standardizationhas become apparent because of the diverse analytical approaches used by individual researchers. In this study, we optimized the image acquisitionconditions for apples, including scaling tool positioning, lighting conditions, and background color selection. In addition, we developed anImageJ-based automated image acquisition and analysis program. We generated 240 images of four apple cultivars (Hongan, Hongro, Fuji, andHwangok) and used 12 image indices to analyze the fruit size, width, length, and shape. We measured the accuracy by comparing the resultswith actual measurements. Significantly high correlation values were observed between fruit width and the major index (R²=0.947-0.993) aswell as between fruit length and the height index (R²=0.964-0.984) based on the analysis using R-squared values to assess accuracy. Thesefindings are expected to enhance the efficiency of apple fruit sorting in the future and can be applied to investigate the shapes of other fruits.