The Open Stamped Parts Dataset (OSPD) provides synthetic and real images of stamped metal sheets for auto manufacturing. It includes unlabeled and labeled images, along with a defect dataset. The dataset replicates the real manufacturing environment, aiding in hole detection models. Machine vision techniques are crucial for part inspection due to dynamic settings. Deep learning approaches are preferred for their ability to learn from data effectively. The dataset aims to enhance defect detection in stamped holes, offering researchers valuable resources for advancing machine vision models.
Ke Bahasa Lain
dari konten sumber
arxiv.org
Wawasan Utama Disaring Dari
by Sara Antiles... pada arxiv.org 03-18-2024
https://arxiv.org/pdf/2403.10369.pdfPertanyaan yang Lebih Dalam