Abstract: With the increasing intelligence of automotive stamping shop and the increasing demand for full process inspection in production, the application of machine vision technology in the field of automotive stamping is also becoming more and more widespread. This article analyzes the working principles of traditional machine vision algorithms and deep learning algorithms, and studies the main application scenarios of machine vision technology in the field of automotive stamping,such as defect detection, feature recognition, etc. The main workflow, algorithms, and advantages in practical application cases are elaborated in detail to promote the promotion and application of machine vision technology in the field of automotive stamping, and look forward to the development prospects of machine vision technology in the field of automotive stamping.

Key words: Machine vision; Stamping field; Defect detection; Feature recognition

摘要： 随着汽车冲压车间的智能化以及生产全流程检测的要求越来越高，机器视觉技术在汽 车冲压领域的应用也越来越广泛。文章通过对传统机器视觉算法及深度学习算法的工作原理 进行剖析，研究了机器视觉技术在汽车冲压领域的主要应用场景，如缺陷检测、特征识别等， 对实际应用案例中的主要工作流程、算法以及优势进行详细阐述，以促进机器视觉技术在汽 车冲压领域的推广应用，并展望机器视觉技术在汽车冲压领域的发展前景。

关键词: 机器视觉；冲压领域；缺陷检测；特征识别