Introduction: The emergence rate of crop seedlings is an important indicator for variety selection, evaluation, field management, and yield prediction. To address the low recognition accuracy caused by the uneven size and varying growth conditions of crop seedlings under salt-alkali stress, this research proposes a peanut seedling recognition model, MS-YOLOv8.Methods: This research employs close-range remote sensing from unmanned aerial vehicles (UAVs) to rapidly recognize and count peanut seedlings. First, a lightweight adaptive feature fusion module (called MSModule) is constructed, which groups the channels of input feature maps and feeds them into different convolutional layers for multi-scale feature extraction. Additionally, the module automatically adjusts the channel weights of each group based on their contribution, improving the feature fusion effect. Second, the neck network structure is reconstructed to enhance recognition capabilities for small objects, and the MPDIoU loss function is introduced to effectively optimize the detection boxes for seedlings with scattered branch growth.Results: Experimental results demonstrate that the proposed MS-YOLOv8 model achieves an AP50 of 97.5% for peanut seedling detection, which is 12.9%, 9.8%, 4.7%, 5.0%, 11.2%, 5.0%, and 3.6% higher than Faster R-CNN, EfficientDet, YOLOv5, YOLOv6, YOLOv7, YOLOv8, and RT-DETR, respectively.Discussion: This research provides valuable insights for crop recognition under extreme environmental stress and lays a theoretical foundation for the development of intelligent production equipment.

Visual harvest maturity is a key visual phenotype for orchard management and harvesting decisions, yet chestnut fruits in natural orchards often exhibit weak color contrast, subtle texture variation, blurred boundaries, and frequent occlusion under complex illumination. This study addresses RGB-based visual harvest maturity recognition and proposes AHM-YOLO, an improved instance segmentation model built upon YOLOv11n-seg. The proposed model enhances maturity-related feature representation by strengthening color- and edge-sensitive cues, stabilizing spatial dependencies under occlusion and illumination variation, and improving cross-scale semantic consistency in dense orchard scenes. A chestnut dataset collected from a typical orchard in Shandong Province is annotated into three visual harvest maturity stages (unripe, semi-ripe, and ripe). To ensure reliable evaluation, the dataset is partitioned at the acquisition unit level, and all experiments are conducted using multi-seed repeated runs. Experimental results show that AHM-YOLO achieves 84.3% Mask mAP50 and 72.2% Mask mAP50–95, demonstrating consistent improvements over the baseline model in complex orchard environments.

The cultivation of saline–alkali-tolerant peanut (Arachis hypogaea L.) varieties can effectively increase grain yield in saline–alkali land. However, traditional assessment methods are often cumbersome and time consuming. To rapidly identify saline–alkali stress-tolerant peanut varieties, this research proposed a saline–alkali stress tolerance evaluation method based on deep learning and multimodal data. Specifically, the research first established multimodal datasets for peanuts at different growth stages and constructed a saline–alkali stress score standard based on unsupervised learning. Subsequently, a deep learning network called BO-MFFNet was built and its structure and hyperparameters were optimized by the Bayes optimization (BO) algorithm. Finally, the point prediction of the saline–alkali stress score were carried out by using the Gaussian process regression model. The experimental results show that the multimodal method is superior to the single-modal data and the BO algorithm significantly improves the performance of the model. The root mean squared error and relative percentage deviation of the BO-MFFNet model are 0.089 and 3.669, respectively. The model effectively predicted the salt–alkali stress tolerance of five varieties, and the predicted results were Huayu25, Yuhua31, Yuhua33, Yuhua32, and Yuhua164 from high to low. This research provides a new method for assessing crop tolerance under extreme environmental stress.

Root-YOLOv7: Multi-scale adaptive object detection and grading of root-knot nematode disease

Accurate estimation of corn ear mass is an essential component of intelligent harvesting. To achieve automated mass measurement of large-scale corn ears, this study developed a dual-view prediction platform and proposed a corn ear mass prediction method based on TS-YOLO and Random Forest. First, a lightweight TS-YOLO instance segmentation model was developed using YOLOv11n-seg as the baseline network to extract morphological features. By incorporating the MStarnet multi-scale feature extraction structure, LSKA attention mechanism, and LADH detection head, the proposed model enhanced feature representation of corn ears and damaged regions while reducing model complexity. Compared with the baseline model, TS-YOLO reduced the number of parameters and computational cost by 15.04% and 11.76%, respectively, and achieved an mAP_@.5 of 0.959. Subsequently, a Random Forest regression model was established by integrating morphological features with corn variety and moisture content information. The Random Forest model yielded an R² of 0.894, with corresponding RMSE and MAE values of 18.19 g and 13.97 g, respectively. Across corn varieties and moisture levels, the mean prediction error was 3.86%, indicating consistent and reliable performance under varying sample conditions. The proposed approach enables automated corn ear mass estimation and provides support for intelligent corn ear inspection.