WaDaBa Dataset
收藏资源简介:
WaDaBa数据集由塞利纳斯大学科学与文学学院等机构创建,旨在通过深度学习技术解决塑料废物分类问题。该数据集包含4000张高分辨率RGB图像,涵盖五种树脂识别码(RIC)类别:PET、PE-HD、PP、PS和其他。数据集通过数据增强技术扩展至11000张图像,以解决类别不平衡问题。数据集的创建过程包括图像采集、标注和数据增强,最终用于训练和测试深度学习模型,特别是YOLO系列模型。该数据集的应用领域主要集中在废物管理和回收,旨在通过自动化技术提高塑料废物分类的效率和准确性。
The WaDaBa dataset was developed by institutions including the Faculty of Science and Literature of Selinus University and other relevant organizations, aiming to solve plastic waste classification problems via deep learning technologies. This dataset contains 4,000 high-resolution RGB images covering five Resin Identification Code (RIC) categories: PET, PE-HD, PP, PS, and others. To address the class imbalance issue, the dataset was expanded to 11,000 images through data augmentation techniques. The creation process of the dataset includes image collection, annotation and data augmentation, and it is ultimately used for training and testing deep learning models, especially YOLO-series models. The application fields of this dataset mainly focus on waste management and recycling, with the goal of improving the efficiency and accuracy of plastic waste classification through automated technologies.

- 1Plastic Waste Classification Using Deep Learning: Insights from the WaDaBa Dataset塞利纳斯大学科学与文学学院, 奥巴费米·阿沃洛沃大学, 伊巴丹大学 · 2024年



