Classification of Soybean [Glycine max (L.) Merr.] Seed Based on Deep Learning Using the YOLOv5 Model

From an agricultural point of view, deep learning models can be used in a variety of way to study the agricultural properties of soybean. Object detection can be performed using image or video data on phenotypic traits of soybean. In this project, a study on the phenotype analysis about soybean seed was conducted by artificial intelligence (AI) based on the YOLOv5 model. In model summary, layers and parameters were calculated as 243 and 7020913, respectively. Means of average precision (mAP)@[0.5: 0.95] was recorded as 0.835, 0.739, 0.785 for each class, and Daewonkong (DW) with yellow seed coat color was calculated as the highest value, and landrace with black seed coat color (NG2) revealed the lowest value. As a result of prediction performance in the confusion matrix, each class of DW, NG2, and inbreeding line with green seed coat color (NGT) showed significant correlation of true positive (TP) in the matrix with the same output value for the input value.