Current crop disease VQA models primarily focus on object counting and detection. However, accurately identifying various disease stages and determining control measures re quire additional knowledge beyond images, including information about control methods and pathogen details. To address this, the VQA dataset relies on the images and questions to retrieve relevant external knowledge. To realize the VQA task of crop diseases external knowledge, we construct the Visual Question Answer Model Based on Crop Diseases External Knowledge for Smart Agriculture (CDEK). CDEK integrates two categories of external knowledge on 66 common dicotyledonous crop diseases by utilizing large language models and agricultural knowledge repositories to enhance knowledge retrieval. This integration enhances the richness of external knowledge repositories. En hancing fine-grained image understanding in CDEK through the utilization of Stack Self-Attention (SSA), utilising Cross Attention and contrastive learning of two external knowledge, with a focus on emphasizing image-related semantic information during training. Finally, an automatic patrol disease detection robot is constructed based on Tensor Processing Unit (TPU) devices and the CDEK model. CDEK achieves an accuracy of 61.7% on the publicly available dataset OK-VQA, surpassing the previous state-of-the-art by 5.1%. Furthermore, we construct the OKiCD-VQA dataset for crop diseases external knowledge and achieve an accuracy of 89.36% using CDEK. A series of ablation experiments are conducted on various modules, the effectiveness of CDEK is demonstrated through extensive experimentation. Contributing solutions to the sustainable development of smart agriculture.

ABSTRACT Pedestrian detection is crucial in agricultural environments to ensure the safe operation of intelligent machinery. In orchards, pedestrians exhibit unpredictable behavior and can pose significant challenges to navigation and operation. This demands reliable detection technologies that ensures safety while addressing the unique challenges of orchard environments, such as dense foliage, uneven terrain, and varying lighting conditions. To address this, we propose ReB‐DINO, a robust and accurate orchard pedestrian detection model based on an improved DINO. Initially, we improve the feature extraction module of DINO using structural re‐parameterization, enhancing accuracy and speed of the model during training and inference decoupling. In addition, a progressive feature fusion module is employed to fuse the extracted features and improve model accuracy. Finally, the network incorporates a convolutional block attention mechanism and an improved loss function to improve pedestrian detection rates. The experimental results demonstrate a 1.6% improvement in Recall on the NREC dataset compared to the baseline. Moreover, the results show a 4.2% improvement in and the number of parameters decreases by 40.2% compared to the original DINO. In the PiFO dataset, the with a threshold of 0.5 reaches 99.4%, demonstrating high detection accuracy in realistic scenarios. Therefore, our model enhances both detection accuracy and real‐time object detection capabilities in apple orchards, maintaining a lightweight attributes, surpassing mainstream object detection models.

Existing pest image classification models require a large number of labeled training images. However, labels for most pest images in the real world do not exist. Therefore, the zero-shot learning method based on generative meta-learning provides an effective solution, which first uses attributes to transfer knowledge from seen classes to unseen classes, and then synthesizes the features of unseen classes. We observe that seen and unseen classes share the same high-level attributes, which can be used to learn a shared set of optimal parameters for seen and unseen classes. Therefore, we propose a novel Hierarchical Class level Attribute guided Generative meta model for pest image Zero-shot Learning (HCAG-ZSL). HCAG-ZSL uses the pre-built Taxonomic Attribute Tree to get the high-level attributes corresponding to the class attributes. These attributes are then fed into a well-designed generator to generate visual features. Extensive experiments show that the proposed model outperforms state-of-the-art generative meta models.