Fischertechnik Industry 4.0 学习工厂对象检测数据集
收藏资源简介:
本数据集是在Fischertechnik Industry 4.0学习工厂环境中创建的,旨在研究不同颜色和外观对对象检测的影响。数据集包含了15种不同材质和颜色的工件,每种工件在相同的环境设置下从32个位置进行拍摄,共收集2475张图片。数据集通过保持摄像头视角、照明条件等环境变量不变,仅改变工件的位置和角度来确保图片的高变异性,用于训练和评估不同大小数据集上的YOLOv8模型,以研究背景特征对对象检测准确性的影响。
This dataset was developed in the Fischertechnik Industry 4.0 learning factory environment, aiming to investigate the effects of different colors and surface appearances on object detection. It contains 15 types of workpieces with distinct materials and colors, and each type was photographed from 32 different positions under identical environmental settings, yielding a total of 2475 images. To ensure high variability in the image collection, environmental variables such as camera perspective and lighting conditions were kept unchanged, while only the position and orientation of the workpieces were modified. This dataset is intended for training and evaluating YOLOv8 models across datasets of varying sizes, to study the influence of background features on object detection accuracy.




