Onion Harvest Detection Dataset

The current agricultural industry faces challenges such as labor shortages and inefficiency, especially during crop harvesting stages. Existing manual detection methods are not only time-consuming but also prone to errors, leading to yield losses. To improve harvesting efficiency and accuracy, this dataset aims to provide high-quality target detection data for smart agriculture systems. The dataset contains images of various onions during the harvesting stage, captured under natural harvesting environments using high-resolution imaging equipment. Through multiple rounds of annotation and expert review, we ensure high data quality and utilize consistency checks to improve annotation accuracy. The data is stored in JPG format, with structured information in JSON files, facilitating subsequent data processing and analysis.