Sunflower image dataset from Bangladesh

The dataset is a comprehensive collection designed for sunflower detection in object detection tasks. The original dataset consists of 1,649 high-quality images capturing various stages and angles of sunflowers. To enhance the model's robustness and generalization capability, data augmentation techniques were applied, expanding the dataset to a final count of 4,286 images. The images are meticulously labeled and presented in the Pascal VOC format, ensuring compatibility with a wide range of deep learning frameworks and facilitating efficient training and evaluation processes in object detection models.

本数据集是为目标检测任务中的向日葵检测场景构建的综合性数据集。原始数据集包含1649张高质量图像，涵盖向日葵的不同生长阶段与拍摄视角。为提升模型的鲁棒性与泛化能力，研究团队通过数据增强技术将数据集扩充至最终的4286张图像。所有图像均经过精细标注，并采用Pascal VOC（帕斯卡视觉对象类挑战赛）格式存储，可兼容绝大多数深度学习框架，便于目标检测模型开展高效的训练与评估流程。