ANÁLISE DE IMAGENS DE PLÂNTULAS PARA A AVALIAÇÃO DO POTENCIAL FISIOLÓGICO DE SEMENTES DE SOJA

Abstract

Image analysis techniques have been studied to evaluate the physical and physiological quality of seeds to minimize the limitations existing in traditional tests. The experiment was carried out in a completely randomized design with four replicates of 50 seeds for all tests used (lot characterization and image analysis). Variance analyses and means compared by ScottKnott test at 5% probability were performed. To compare the accelerated aging, seedling length and root tests with image analysis, a factor analysis was performed, and the factors considered were lots and types of tests (6 x 2) for each variety. The aim of this study was to verify the efficiency of GroundEye ® equipment in to evaluatesoybean seeds vigor from seedling images ingermination test, its possible application as a complementary test in the quality control of seed producing companies. The equipment proved to be efficient for its use as a technique for analyzing seedling images, becoming an internal audit, auditing this by artificial intelligence, thus eliminating any human interference that may have when checking the seed physiological potential.