Non-Contact Dimensional Quality Inspection System of Prefabricated Components Using 3D Matrix Camera

Dimensional quality inspection of prefabricated components is crucial for ensuring building quality and safety. Currently, manual measurement methods are predominantly used in dimensional quality inspection of prefabricated components, which are both time-consuming and labor-intensive, constraining production efficiency. This study thus developed a non-contact image measurement system using an innovative three-dimensional (3D) matrix camera, which automatically performed dimensional quality inspection, utilizing technologies such as a parallel optical axis four-camera matrix imaging and machine learning algorithms. Compared to traditional techniques, this system exhibited enhanced adaptability to the manufacturing process of prefabricated components, along with desirable accuracy and efficiency. Building upon a comprehensive literature review, the hardware constituents of the 3D matrix camera image measurement system were meticulously introduced, followed by the underlying principles and implementations of data acquisition, processing and comparison methods, including parallel optical axis four-camera matrix imaging, automatic stitching algorithms for 3D point clouds, feature recognition algorithms, and matching principles. The feasibility of the proposed system was validated through a case study analysis. The application results indicated that the system was capable of automatically performing non-contact measurements of dimensional deviations in prefabricated components with an accuracy of ±3 mm, thereby enhancing production quality.