合成热图数据集
收藏资源简介:
合成热图数据集是由印度孟买理工学院机械工程系创建的,用于钢板裂纹检测的深度学习研究。该数据集通过有限元模拟生成,包含100张图像,每张图像都经过标注以标记裂纹位置。数据集的创建过程结合了有限元模拟和数据增强技术,旨在提高深度学习模型在裂纹检测中的性能。该数据集主要应用于计算机视觉领域,特别是自动化裂纹检测,旨在减少人工干预并提高检测效率。
The Synthetic Heatmap Dataset was developed by the Department of Mechanical Engineering, Indian Institute of Technology Bombay, for deep learning research focused on steel plate crack detection. Generated through finite element simulation, this dataset comprises 100 images, each annotated with the exact positions of cracks. The dataset creation process integrates finite element simulation and data augmentation technologies, with the objective of improving the performance of deep learning models for crack detection. Primarily applied in the computer vision domain, especially for automated crack detection, this dataset aims to reduce manual intervention and enhance detection efficiency.




