Green Pepper Quantity Statistics Dataset

The current agricultural industry faces challenges of low efficiency in crop management and inaccurate yield predictions, especially in the cultivation and management of green peppers, due to the lack of effective monitoring methods. Existing solutions largely rely on manual observation, which is inefficient and prone to errors, falling short of modern agricultural development needs. This dataset aims to provide high-quality green pepper target detection data, helping researchers and developers establish more precise crop monitoring systems. Data collection is carried out using high-resolution cameras in a greenhouse environment, ensuring consistent light and temperature conditions, greatly increasing the data's usability. To ensure data quality, a multi-round annotation and expert review mechanism is employed to ensure annotation consistency and accuracy. Data is stored in JPG format, organized by category and date for ease of subsequent use and management. The core advantage of this dataset lies in its high annotation precision, with annotation consistency exceeding 95% and data completeness over 90%. Additionally, a new data augmentation technique is introduced, using Generative Adversarial Networks (GAN) to enhance the model's generalization capability, resulting in a 15% improvement in performance metrics for target detection. This dataset not only provides a solid data foundation for smart agriculture development but also offers a practical solution to green pepper monitoring and management challenges.