Lang Qiao

High-resolution UAV photogrammetry has become a key technology for precision agriculture, enabling centimeter-level crop monitoring and point-level plant localization. However, point-level maize localization in UAV imagery remains challenging due to (1) extremely small object-to-pixel ratios, typically less than 0.1%, (2) prohibitive computational costs of quadratic attention on ultra-high-resolution images larger than 3000 x 4000 pixels, and (3) agricultural scene-specific complexities such as sparse object distribution and environmental variability that are poorly handled by general-purpose vision models. To address these challenges, we propose the Additive Kolmogorov-Arnold Transformer (AKT), which replaces conventional multilayer perceptrons with Pade Kolmogorov-Arnold Network (PKAN) modules to enhance functional expressivity for small-object feature extraction, and introduces PKAN Additive Attention (PAA) to model multiscale spatial dependencies with reduced computational complexity. In addition, we present the Point-based Maize Localization (PML) dataset, consisting of 1,928 high-resolution UAV images with approximately 501,000 point annotations collected under real field conditions. Extensive experiments show that AKT achieves an average F1-score of 62.8%, outperforming state-of-the-art methods by 4.2%, while reducing FLOPs by 12.6% and improving inference throughput by 20.7%. For downstream tasks, AKT attains a mean absolute error of 7.1 in stand counting and a root mean square error of 1.95-1.97 cm in interplant spacing estimation. These results demonstrate that integrating Kolmogorov-Arnold representation theory with efficient attention mechanisms offers an effective framework for high-resolution agricultural remote sensing.

The rapid development of unmanned aerial vehicles (UAVs) and deep learning has accelerated the application of semantic segmentation in precision agriculture (SSPA). A key driver of this progress lies in multimodal fusion, which leverages complementary structural, spectral, and physiological information to enhance the representation of complex agricultural scenes. Despite advancements, the efficacy of multimodal fusion in SSPA is limited by modality heterogeneity and the difficulty of simultaneously retaining fine details and capturing global context. To address these challenges, we propose AgriFusion, a dual-encoder framework based on convolutional and transformer architectures for SSPA tasks. Specifically, convolutional and transformer encoders are first used to extract crop-related local structural details and global contextual features from multimodal inputs. Then, an attention-based fusion module adaptively integrates these complementary features in a modality-aware manner. Finally, a MLP-based decoder aggregates multi-scale representations to generate accurate segmentation results efficiently. Experiments conducted on the Agriculture-Vision dataset demonstrate that AgriFusion achieves a mean Intersection over Union (mIoU) of 49.31%, Pixel Accuracy (PA) of 81.72%, and F1 score of 67.85%, outperforming competitive baselines including SegFormer, DeepLab, and AAFormer. Ablation studies further reveal that unimodal or shallow fusion strategies suffer from limited discriminative capacity, whereas AgriFusion adaptively integrates complementary multimodal features and balances fine-grained local detail with global contextual information, yielding consistent improvements in identifying planting anomalies and crop stresses. These findings validate our central claims that modality-aware spectral fusion and balanced multi-scale representation are critical to advancing agricultural semantic segmentation, and establish AgriFusion as a principled framework for enhancing remote sensing-based monitoring with practical implications for sustainable crop management and precision farming.

In order to improve the diagnosis accuracy of chlorophyll content in maize canopy, the remote sensing image of maize canopy with multiple growth stages was acquired by using an unmanned aerial vehicle (UAV) equipped with a spectral camera. The dynamic influencing factors of the canopy multispectral images of maize were removed by using different image segmentation methods. The chlorophyll content of maize in the field was diagnosed. The crop canopy spectral reflectance, coverage, and texture information are combined to discuss the different segmentation methods. A full-grown maize canopy chlorophyll content diagnostic model was created on the basis of the different segmentation methods. Results showed that different segmentation methods have variations in the extraction of maize canopy parameters. The wavelet segmentation method demonstrated better advantages than threshold and ExG index segmentation methods. This method segments the soil background, reduces the texture complexity of the image, and achieves satisfactory results. The maize canopy multispectral band reflectance and vegetation index were extracted on the basis of the different segmentation methods. A partial least square regression algorithm was used to construct a full-grown maize canopy chlorophyll content diagnostic model. The result showed that the model accuracy was low when the image background was not removed (Rc² (the determination coefficient of calibration set) = 0.5431, RMSEF (the root mean squared error of forecast) = 4.2184, MAE (the mean absolute error) = 3.24; Rv² (the determination coefficient of validation set) = 0.5894, RMSEP (the root mean squared error of prediction) = 4.6947, and MAE = 3.36). The diagnostic accuracy of the chlorophyll content could be improved by extracting the maize canopy through the segmentation method, which was based on the wavelet segmentation method. The maize canopy chlorophyll content diagnostic model had the highest accuracy (Rc² = 0.6638, RMSEF = 3.6211, MAE = 2.89; Rv² = 0.6923, RMSEP = 3.9067, and MAE = 3.19). The research can provide a feasible method for crop growth and nutrition monitoring on the basis of the UAV platform and has a guiding significance for crop cultivation management.