A new image recognition algorithm, YOLOv8n-WSE-pest, has been developed to enhance pest management in tea plantations in Yunnan Province, China.

This algorithm, bolstered by techniques such as structural pruning and depthwise separable convolution, enables real-time detection of strawberry pests and diseases, addressing challenges related to speed, accuracy, and computational load.

The research indicates that the YOLOv8n-WSE-pest model significantly improves the efficiency and accuracy of pest management practices.

The urgency of such advancements is underscored by global food security threats posed by population growth, disease, resource limitations, and climate change.

The advancements presented in this study contribute significantly to the development of smart agricultural technologies, offering new tools for effective disease detection.

Experimental results demonstrate that the CUIB-YOLO model effectively balances computational efficiency with detection accuracy, making it suitable for devices with limited resources.

The methodology employed includes dataset preparation, CNN model training, deployment on edge devices, and performance evaluation to ensure practicality in real-world scenarios.

Future research aims to further enhance detection accuracy, particularly for small target images, while maintaining the lightweight properties of the model.

Challenges remain in the practical implementation of these technologies, including the need for large annotated datasets and model interpretability.

The study will detail related work, methodology, experimental results, and conclusions regarding the proposed models in future sections.

Traditional disease detection methods are often labor-intensive and prone to human error, highlighting the need for more scalable and accurate solutions.

Integrating machine learning algorithms with imaging data enhances spatial-temporal disease identification, aiding in timely detection and management.