Sashuang Sun

The phenotypic analysis of wheat spikes plays an important role in wheat growth management, plant breeding, and yield estimation. However, the dense and tight arrangement of spikelets and grains on the spikes makes the phenotyping more challenging. This study proposed a rapid and accurate image-based method for in-field wheat spike phenotyping consisting of three steps: wheat spikelet segmentation, grain number classification, and total grain number counting. Wheat samples ranging from the early filling period to the mature period were involved in the study, including three varieties: Zhengmai 618, Yannong 19, and Sumai 8. In the first step, the in-field collected images of wheat spikes were optimized by perspective transformation, augmentation, and size reduction. The YOLOv8-seg instance segmentation model was used to segment spikelets from wheat spike images. In the second step, the number of grains in each spikelet was classified by a machine learning model like the Support Vector Machine (SVM) model, utilizing 52 image features extracted for each spikelet, involving shape, color, and texture features as the input. Finally, the total number of grains on each wheat spike was counted by adding the number of grains in the corresponding spikelets. The results showed that the YOLOv8-seg model achieved excellent segmentation performance, with an average precision (AP) @[0.50:0.95] and accuracy (A) of 0.858 and 100%. Meanwhile, the SVM model had good classification performance for the number of grains in spikelets, and the accuracy, precision, recall, and F1 score reached 0.855, 0.860, 0.865, and 0.863, respectively. Mean absolute error (MAE) and mean absolute percentage error (MAPE) were as low as 1.04 and 5% when counting the total number of grains in the frontal view wheat spike images. The proposed method meets the practical application requirements of obtaining trait parameters of wheat spikes and contributes to intelligent and non-destructive spike phenotyping.

The estimation accuracy of plant dry matter by spectra- or remote sensing-based methods tends to decline when canopy coverage approaches closure; this is known as the saturation problem. This study aimed to enhance the estimation accuracy of plant dry matter and subsequently use the critical nitrogen dilution curve (CNDC) to diagnose N in Choy Sum by analyzing the combined information of canopy imaging and plant height. A three-year experiment with different N levels (0, 25, 50, 100, 150, and 200 kg∙ha⁻¹) was conducted on Choy Sum. Variables of canopy coverage (CC) and plant height were used to build the dry matter and N estimation model. The results showed that the yields of N₀ and N₂₅ were significantly lower than those of high-N treatments (N₅₀, N₁₀₀, N₁₅₀, and N₂₀₀) for all three years. The variables of CC × Height had a significant linear relationship with dry matter, with R² values above 0.87. The good performance of the CC × Height-based model implied that the saturation problem of dry matter prediction was well-addressed. By contrast, the relationship between dry matter and CC was best fitted by an exponential function. CNDC models built based on CC × Height information could satisfactorily differentiate groups of N deficiency and N abundance treatments, implying their feasibility in diagnosing N status. N application rates of 50–100 kgN/ha are recommended as optimal for a good yield of Choy Sum production in the study region.