Seedling Health Status Recognition Dataset

The current agricultural field faces challenges such as the difficulty in monitoring crop health and the untimely identification of diseases, especially during the seedling stage, often resulting in significant economic losses due to the lack of timely data support. Existing agricultural monitoring methods largely rely on manual inspection, which is low in efficiency and insufficient in accuracy. This dataset aims to support the research and application of object detection technology by providing high-quality annotated images of seedling health status. Data collection is conducted by drones in greenhouse environments, combined with high-resolution cameras to capture images of seedling growth status in real-time. For quality control, a combination of multiple rounds of annotation and expert reviews is used to ensure the accuracy and consistency of the annotations. The data storage format is JPEG, organized by the correspondence between image folders and annotation files. The core advantage of this dataset is its high annotation accuracy, with an average annotation consistency exceeding 95%. By introducing deep learning techniques for data enhancement, the model's generalization capability is improved. After applying it to crop health monitoring, detection accuracy increased by 20%, effectively reducing crop disease losses.