Hao Zhou

Several pressing issues have been revealed by deep learning-based tomato ripening detection technology in intricate environmental applications: The ripening transition stage distinction is not accurate enough, small target tomato detection is likely to miss, and the detection technology is more susceptible to variations in light. Based on the YOLOv11 model, a YOLOv11-SLBA tomato ripeness detection model was presented in this study. First, SPPF-LSKA is used in place of SPPF in the backbone section, greatly improving the model’s feature discrimination performance in challenging scenarios including dense occlusion and uneven illumination. Second, a new BiAttFPN hierarchical progressive fusion is added in the neck area to increase the feature retention of small targets during occlusion. Lastly, the feature separability of comparable categories is significantly enhanced by the addition of the auxiliary detection head DetectAux. In this study, comparative experiments are carried out to confirm the model performance. Under identical settings, the YOLOv11-SLBA model is compared to other target detection networks, including Faster R-CNN, SSD, RT-DETR, YOLOv7, YOLOv8, and YOLOv11. With 2.7 million parameters and 10.9 MB of model memory, the YOLOv11-SLBA model achieves 92% P, 83.5% R, 91.3% mAP50, 64.6% mAP50-95, and 87.5% F1-score. This is a 3.4% improvement in accuracy, a 1.5% improvement in average precision, and a 1.6% improvement in F1-score when compared to the baseline model YOLOv11. It outperformed the other comparison models in every indication and saw a 1.6% improvement in score. Furthermore, the tomato-ripeness1public dataset was used to test the YOLOv11-SLBA model, yielding model p values of 78.6%, R values of 91.5%, mAP50 values of 93.7%, and F1-scores of 84.6%. This demonstrates that the model can perform well across a variety of datasets, greatly enhances the detection generalization capability in intricate settings, and serves as a guide for the algorithm design of the picking robot vision system.

Chlorophyll content is an important indicator for estimating potato growth. However, there are still some research gaps in the inversion of canopy chlorophyll content using unmanned aerial vehicle (UAV) remote sensing. For example, it faces limitations of the growth cycle, low parameter accuracy, and single feature selection, and there is a lack of efficient and precise systematic research methods. In this study, an improved Adaptive-Forward Feature Selection (AFFS) algorithm was developed by combining remote sensing data and measured data to optimize the input Vegetation Index (VI) variables. Gradient Boosting Machine (GBM) model parameters were optimized using a hybrid strategy improved Elephant Herd Optimization (EHO) algorithm (CDE-EHO) that combines Differential Evolution (DE) and Cauchy Mutation (CM). The CDE-EHO method optimizes the GBM model, achieving maximum accuracy, according to the testing results. The optimal coefficients of determination (R²) values of the prediction set are 0.663, 0.683, and 0.906, respectively, the Root Mean Squared Error (RMSE) values are 2.673, 3.218, and 2.480, respectively, and the Mean Absolute Error (MAE) values are 2.052, 2.732, and 1.928, respectively, during the seedling stage, tuber expansion stage and cross-growth stage. This approach has significantly enhanced the inversion model’s prediction performance as compared to earlier research. The chlorophyll content in the potato canopy has been accurately extracted in this work, offering fresh perspectives and sources for further research in this area.

Potato is a major food crop in China. Its development and nutritional state can be inferred by the content of chlorophyll in its canopy. However, the existing study on applying feature extraction and optimization algorithms to determine the canopy SPAD (Soil–Plant Analytical Development) values of potatoes at various fertility stages is inadequate and not very reliable. Using the Pearson feature selection algorithm and the Competitive Adaptive Reweighted Sampling (CARS) method, the Vegetation Index (VI) with the highest correlation was selected as a training feature depended on multispectral orthophoto images from unmanned aerial vehicle (UAV) and measured SPAD values. At various potato fertility stages, Random Forest (RF), Support Vector Regression (SVR), and Extreme Gradient Boosting (XGBoost) inversion models were constructed. The models’ parameters were then optimized using the Grey Wolf Optimizer (GWO) and Sparrow Search Algorithm (SSA). The findings demonstrated a higher correlation between the feature selected VI and SPAD values; additionally, the optimization algorithm enhanced the models’ prediction accuracy; finally, the addition of the fertility stage feature considerably increased the accuracy of the full fertility stage in comparison to the single fertility stage. The models with the highest inversion accuracy were the CARS-SSA-RF, CARS-SSA-XGBoost, and Pearson-SSA-XGBoost models. For the single-fertility and full-fertility phases, respectively, the optimal coefficients of determination (R²s) were 0.60, 0.66, and 0.87, the root-mean-square errors (RMSEs) were 2.63, 3.23, and 2.39, and the mean absolute errors (MAEs) were 2.00, 2.75, and 1.99.

Potato, a vital food and cash crop, necessitates precise identification and area estimation for effective planting planning, market regulation, and yield forecasting. However, extracting large-scale crop areas using satellite remote sensing is fraught with challenges, such as low spatial resolution, cloud interference, and revisit cycle limitations, impeding the creation of high-quality time–series datasets. In this study, we developed a high-resolution vegetation index time–series by calculating coordination coefficients and integrating reflectance data from Landsat-8, Landsat-9, and Sentinel-2 satellites. The vegetation index time–series were enhanced through using linear interpolation and Savitzky–Golay (S-G) filtering to reconstruct high-quality data. We employed the harmonic analysis of NDVI time–series (HANTS) method to extract features from the time–series and evaluated the classification accuracy across five feature sets: vegetation index time–series features, band means, vegetation index means, texture features, and color space features. The Random Forest (RF) model, utilizing the full feature set, emerged as the most accurate, achieving a precision rate of 0.97 and a kappa value of 0.94. We further refined the feature subset using the SHAP-SFS feature selection method, leading to the SHAP-SFS-RF classification approach for differentiating potato from non-potato crops. This approach enhanced accuracy by approximately 0.1 and kappa value by around 0.2 compared to the RF model, with the extracted areas closely aligning with statistical yearbook data. Our study successfully achieved the accurate extraction of potato planting areas at the county level, offering novel insights and methodologies for related research fields.