Crop Leaf Anomaly Detection Dataset

The current agricultural sector faces crop yield reduction issues due to pests and diseases, particularly with insufficient monitoring of leaf abnormalities, causing significant losses to farmers. Existing monitoring methods largely rely on manual inspection, which is inefficient and prone to omissions, necessitating automated detection solutions. This dataset aims to provide AI algorithms with high-quality leaf anomaly samples to address the issue of insufficient detection accuracy. Data collection is primarily conducted using high-resolution cameras under natural conditions, covering various anomalies such as insect holes, cracks, and dryness. We use multiple rounds of annotation and expert review to ensure data quality. All data is stored in JPG format, accompanied by JSON files that record annotation information. The dataset's organization is clear, facilitating subsequent model training and performance evaluation. The dataset's core advantages lie in its high annotation accuracy and consistency, achieving over 95% annotation consistency, greatly reducing the risk of misjudgments and omissions. Additionally, utilizing newly developed image enhancement technology, the model's detection accuracy has increased by 15%, significantly improving the efficiency and accuracy of crop health assessment.