Crop Pest Region Segmentation Dataset

The current agricultural sector faces the challenge of frequent pest outbreaks, which not only affect crop yield but also reduce the quality of agricultural products. Existing monitoring methods largely rely on manual inspection, which is inefficient and prone to errors. To improve the accuracy and efficiency of pest monitoring, this dataset offers high-quality pixel-level segmentation annotations of leaf pest regions, aimed at supporting the training and optimization of AI models. The dataset was collected by photographing crop leaves in real farmland with high-resolution cameras to ensure image clarity and detail. Multiple rounds of annotation and consistency checks were employed throughout the annotation process to guarantee data quality and accuracy. The data is stored in PNG/JPG format, with each image corresponding to a segmentation mask, facilitating model training and validation.