Pumpkin Counting Dataset

The current agricultural industry faces challenges of low crop monitoring efficiency and significant human counting errors. Existing solutions often rely on manual experience, lacking automation and accuracy. This dataset aims to provide a high-quality object detection training foundation for agricultural image recognition, meeting the need for automated pumpkin counting. The dataset collects pumpkin images in various environments using professional equipment, and undergoes multiple rounds of annotation and consistency checks to ensure high data quality. The data is stored in JPG format, with annotation information organized in JSON format, facilitating subsequent model training and evaluation.