BDWaste: A comprehensive image dataset of digestible and indigestible waste in Bangladesh

The dataset describes about two significant classes of waste, which are defined as “waste management." Two classes include biodegradable and non-biodegradable, which contain a total of 21 distinct classes of waste. Biodegradable contains potato peel, lemon peel, rice, paper, fish scale, malt shell, banana peel, coffee cup, sugarcane husk, and egg shell. Non-biodegradable materials include polythene, plastic, wire, glass, gloves, chip packets, empty medicine packets, masks, and bottles. The dataset holds an overall total of 2625 real images of waste. Each image, both outdoor and indoor, used natural light to avoid shadows with an appropriate background. The device used to capture the images is an Android. The whole dataset practices the primary concepts of waste management, gives an overview of improper waste management, provides guidance on recycling, and also serves as valuable resources on waste management for future development purposes.本数据集聚焦两类归属于“废物管理”研究范畴的典型废弃物。该两大类别为可降解（biodegradable）与不可降解（non-biodegradable）废弃物，其下总计涵盖21种不同的废弃物细分品类。其中可降解类废弃物包含土豆皮、柠檬皮、大米、纸张、鱼鳞、麦芽壳、香蕉皮、咖啡杯、甘蔗渣以及蛋壳；不可降解类废弃物包含聚乙烯（polythene）制品、塑料、金属丝、玻璃、手套、薯片包装袋、空药盒、口罩以及各类瓶子。本数据集总计包含2625张真实的废弃物实拍图像，所有图像均在室内或室外场景下拍摄，采用自然光以避免产生阴影，并搭配了适宜的背景环境。图像采集设备为安卓（Android）设备。本数据集涵盖了废物管理的核心要义，梳理了不当废物管理的典型问题，为废弃物回收利用提供了科学指导，同时也可为未来废物管理领域的研发工作提供极具价值的参考资源。