3D Shape Measurements by Images
摘要
In this chapter, we present the fundamental principles behind the most used classical and modern techniques for measuring the 3D shape of objects or surrounding scenes using 2D image processing and structured light. These principles form the basis of what are widely known as optical “scanners”. Over the past few decades, these systems have significantly advanced in terms of measurement range (from small to very large objects), resolution, and measurement uncertainty. As a result, they have enabled a wide array of applications across various sectors. These include industrial use cases such as reverse engineering, quality assurance, and tolerance control of products and mechanical components. They are also applied to large-scale measurements, such as capturing the geometry of entire cars, ships, aircraft, and their parts; mapping buildings, bridges, and natural landscapes; and documenting cultural heritage through the 3D digitization of artworks, statues, monuments, and historical sites, for purposes of visualization, conservation, and restoration. The chapter discusses both active techniques (which use structured light) and passive techniques (which rely on natural illumination), alongside key image processing methods. These techniques enable the extraction of (x, y, z) coordinates for millions of surface points, collectively referred to as a point cloud. This point cloud forms the foundation for creating mathematical surface models and geometric primitives, measuring dimensional properties, comparing against CAD models, developing new models, or replicating objects through 3D printing, including scaled versions.