6D object pose estimation is an important research topic in computer vision and robotics, and has been widely applied in industrial automation, robotic manipulation, and augmented reality. In view of the rapid development of 6D object pose estimation in recent years, this paper presents a systematic survey of related research progress. First, existing methods are categorized into traditional methods and deep learning-based methods. The deep learning approaches are further divided into two-stage methods and single-stage methods. The basic ideas, key techniques, and characteristics of these methods are analyzed and compared. Then, commonly used public datasets and evaluation metrics, including ADD, ADD-S, VSD, and AR, are summarized. The performance of representative algorithms on public datasets is also reviewed and analyzed. Based on these analyses, the performance of different methods under complex scenarios, occlusion conditions, and real-time requirements is discussed. Finally, considering practical application demands such as industrial automation and robotic manipulation, the challenges and future research directions of 6D object pose estimation are summarized and discussed. This survey aims to provide a comprehensive reference for researchers and practitioners in understanding the development of this field and selecting appropriate methods for practical applications.