Currently, the demand for refined crop monitoring through remote sensing is increasing rapidly. Due to the similar spectral and morphological characteristics of different crops and vegetation, traditional methods often rely on deeper neural networks to extract meaningful features. However, deeper networks face a key challenge: while extracting deep features, they often lose some boundary details and small-plot characteristics, leading to inaccurate farmland boundary classifications. To address this issue, we propose the Detail and Deep Feature Multi-Branch Fusion Network for High-Resolution Farmland Remote-Sensing Segmentation (DFBNet). DFBNet introduces an new three-branch structure based on the traditional UNet. This structure enhances the detail of ground objects, deep features across multiple scales, and boundary features. As a result, DFBNet effectively preserves the overall characteristics of farmland plots while retaining fine-grained ground object details and ensuring boundary continuity. In our experiments, DFBNet was compared with five traditional methods and demonstrated significant improvements in overall accuracy and boundary segmentation. On the Hi-CNA dataset, DFBNet achieved 88.34% accuracy, 89.41% pixel accuracy, and an IoU of 78.75%. On the Netherlands Agricultural Land Dataset, it achieved 90.63% accuracy, 91.6% pixel accuracy, and an IoU of 83.67%. These results highlight DFBNet’s ability to accurately delineate farmland boundaries, offering robust support for agricultural yield estimation and precision farming decision-making.

Remote sensing image segmentation holds significant application value in precision agriculture, environmental monitoring, and other fields. However, in the task of fine-grained segmentation of weeds and crops, traditional deep learning methods often fail to balance global semantic information with local detail features, resulting in over-segmentation or under-segmentation issues. To address this challenge, this paper proposes a segmentation model based on a dual-branch Transformer architecture—DBFormer—to enhance the accuracy of weed detection in remote sensing images. This approach integrates the following techniques: (1) a dynamic context aggregation branch (DCA-Branch) with adaptive downsampling attention to model long-range dependencies and suppress background noise, and (2) a local detail enhancement branch (LDE-Branch) leveraging depthwise-separable convolutions with residual refinement to preserve and sharpen small weed edges. An Edge-Aware Loss module further reinforces boundary clarity. On the Tobacco Dataset, DBFormer achieves an mIoU of 86.48%, outperforming the best baseline by 3.83%; on the Sunflower Dataset, it reaches 85.49% mIoU, a 4.43% absolute gain. These results demonstrate that our dual-branch synergy effectively resolves the global–local conflict, delivering superior accuracy and stability in the context of practical agricultural applications.

The complexity of data and business content in enterprise software design makes it difficult to grasp, which leads to problems such as low team coding efficiency and easy errors. The key to solving this problem is to take the business data of the enterprise as the center and drive the code development with the characteristics of the data itself. Therefore, this study proposes an automatic front-end code generation method based on data and template integration (GC-DT). This method utilizes the enterprise business table structure and templates to form the process of automatic code generation, and realizes the automatic generation of VUE front-end code. Practical application shows that compared with the traditional method, this method can effectively reduce the workload of software development, shorten the software development cycle, and still effectively maintain the consistency and correctness of the team code in the face of complex content structure. Therefore, this method has high application value in enterprise software development.

Large cross-modal models such as contrastive language-image pre-training suffer from heavy computation costs and weak anti-interference ability, which brings obstacles to lightweight deployment on edge devices with limited computing resources, a common requirement in industrial edge devices. Traditional spatial visual prompt learning cannot mine discriminative frequency texture features, resulting in poor generalization under scarce labeled data and noisy industrial scenes. This paper proposes frequency-domain adaptive visual prompt learning (FD-AVPL). The framework decomposes images into multi-band frequency features, introduces weighted attention fusion, and generates frequency-aware prompts to fine-tune pre-trained models with tiny parameter overhead. Tests on five public datasets show FD-AVPL outperforms state-of-the-art prompt methods by 2.0%–9.0% accuracy in full-shot and few-shot scenarios. It maintains strong robustness to noise, occlusion and illumination shifts, adding only 1.1 M/1.3 M trainable parameters for RN50/ViT-B/32. The proposed method provides a lightweight parameter-efficient tuning scheme for large vision-language models, which is well-suited to the deployment demands of lightweight tuning scheme for deploying large vision-language models on resource-limited embedded terminals including industrial terminals.