3D reconstruction of energy meter PCB based on highly similar keyframe screening

With the advancement of technologies such as the Internet of Things (IOT) and artificial intelligence, real-time monitoring and control of various stages of product production and operation through digital twin systems have become a new trend in the manufacturing industry. Three-dimensional reconstruction is a key technology of digital twin systems. However, in the context of intelligent manufacturing for energy meters, the multitude and diversity of components on the PCB pose challenges for target identification, thereby complicating the three-dimensional reconstruction process. This paper presents an optimized keyframe screening algorithm based on YOLO recognition, which extracts highly similar key frames from videos to identify the positional framework of circuit components in energy meters. Additionally, it preprocesses image brightness and clarity to enhance the recognition accuracy of the YOLO algorithm, with the identified component information being utilized for three-dimensional reconstruction. The virtual models constructed using this method not only achieve high efficiency in component detection but also provide reliable data support for circuit design. By enabling precise component identification and high-quality virtual model construction, the proposed method demonstrates significant application potential in the fields of intelligent manufacturing, industrial automation, and virtual reality.