Automatic and Accurate Acquisition of Stem-Related Phenotypes of Mature Soybean Based on Deep Learning and Directed Search Algorithms

Abstract

The stem-related phenotype of mature stage soybean is important in soybean material selection. How to improve on traditional manual methods and obtain the stem-related phenotype of soybean more quickly and accurately is a problem faced by producers. With the development of smart agriculture, many scientists have explored soybean phenotypes and proposed new acquisition methods, but soybean mature stem-related phenotype studies are relatively scarce. In this study, we used a deep learning method within the convolutional neural network to detect mature soybean stem nodes and identified soybean structural features through a novel directed search algorithm. We subsequently obtained the pitch number, internodal length, branch number, branching angle, plant type spatial conformation, plant height, main stem length, and new phenotype-stem curvature. After 300 epochs, we compared the recognition results of various detection algorithms to select the best. Among them, YOLOX had a maximum average accuracy (mAP) of 94.36% for soybean stem nodes and scale markers. Through comparison of the phenotypic information extracted by the directed search algorithm with the manual measurement results, we obtained the Pearson correlation coefficients, R, of plant height, pitch number, internodal length, main stem length, stem curvature, and branching angle, which were 0.9904, 0.9853, 0.9861, 0.9925, 0.9084, and 0.9391, respectively. These results show that our algorithm can be used for robust measurements and counting of soybean phenotype information, which can reduce labor intensity, improve efficiency, and accelerate soybean breeding.

Deep Learning for Soybean Monitoring and Management

“…It is important to emphasize, however, that the high classification accuracies are almost always obtained on data with rather limited variability, so the reported results are not necessarily valid for real practical conditions (this issue is addressed in more depth in the next section). Disease recognition RGB DenseNet201 0.97 1 [23] Water stress detection RGB AlexNet, GoogLeNet, Inception V3 0.93 1 [24] Root phenotyping RGB CAE 0.66-0.99 [15] Weed detection RGB JULE, DeepCluster 0.97 1 [11] Volunteer corn detection RGB, CIR GoogleNet 0.99 1 [25] Disease severity RGB FPN, U-Net, DeepLabv3+ 0.95-0.98 [3] Pest detection HS Attention-ResNet 0.95 1 [26] Stem phenotyping RGB YOLO X 0.94 1 [27] Pod detection, yield prediction RGB YOLO v5 0.94 4 [28] Disease recognition RGB DCNN 0.98 1 [29] Seed counting RGB Two-column CNN 0.82-0.94 [30] Pest detection RGB Modified YOLO v4 0.87 2 [31] Disease severity RGB RetinaNet 0.64-0.65 1,2 [32] Weed detection RGB Faster R-CNN, YOLO v3 0.89-0.98 [19] Defoliation estimation RGB, synthetic AlexNet, VGGNet and ResNet 0.98 3 [33] Disease recognition RGB DIM-U-Net, SR-AE, LSTM 0.99 2 [34] Weed detection RGB DCNN 0.93 1 [35] Pest detection RGB Several CNNs 0.94 1 [12] Seed-per-pot estimation RGB DCNN 0.86 1 [36] Cultivar identification RGB ResNet-50, DenseNet-121, DenseNet 0.84 1 [37] Disease recognition RGB AlexNet, GoogLeNet, ResNet-50 0.94 1 [38] Pod counting RGB YOLO POD 0.97 4 [39] Seed phenotyping RGB, synthetic Mask R-CNN 0.84-0.90 [40] Yield prediction, biomass HS DCNN 0.76-0.91 [41] Disease recognition RGB GAN 0.96 1 [42] Disease recognition RGB Faster R-CNN 0.83 5 [43] Weed detection RGB Faster R-CNN 0.99 1 [44] Pod counting RGB R-CNN, YOLO v3, YOLO v4, YOLO X 0.90-0.98 [45] Seed defect recognition RGB MobileNet V2 0.98 1 [46] Seed counting RGB P2PNet-Soy 0.87 4 [47] Cultivar identification HS DCNN 0.90 1 …”

Study on Utilizing Mask R-CNN for Phenotypic Estimation of Lettuce’s Growth Status and Optimal Harvest Timing

“…In recent years, with the development of computer vision and deep learning techniques, image-based methods for measuring plant phenotypic traits have received widespread attention and application. Image sensors collect image information of plants, such as RGB cameras [5][6][7][8][9], depth cameras [10][11][12][13], and optical spectrometers [14,15], which are then analyzed and processed with deep learning models to facilitate the automated, precise, and intelligent measurement of plant phenotypic traits. Li et al [16] used UAV multispectral data as input to predict winter wheat yield, the convolutional neural network (CNN) model gave the best results, with R 2 of 0.752 and NMSE of 0.404 t•ha −1 .…”

“…The results demonstrated that transfer learning based on the VGG19 model achieved the highest accuracy for both crops. Guo et al [5] used a deep learning method coupled with a novel directed search algorithm to obtain stem-related phenotypes for soybeans. The Pearson correlation coefficients (R) of plant height, pitch number, internodal length, main stem length, stem curvature, and branching angle were 0.9904, 0.9853, 0.9861, 0.9925, 0.9084, and 0.9391, respectively.…”