The display, comparison, and registration of surfaces in range data format

Geometric Modeling involves the pictorial synthesis of an object from an associated computer-based model. It involves defining a geometric model and developing design methods for this model. Modeling the surface is an important aspect of geometric modeling, where an object is represented by its outer skin (boundary) as discrete elements, and these elements are not associated with the architectural structure of the object itself.; Surface registration is the process of aligning surfaces represented in the form of a digital image, and has applications in image processing and analysis. Without objective criteria for measuring the likeness between surfaces, the reliability of surface registration is difficult to assess. This thesis defines different strategies for measuring the likeness between surfaces, so that the comparison and registration of the surfaces can be achieved. The measure of likeness between two digital surfaces is quantified by a Surface Similarity Measure (SSM) and the data representation and display methods for surfaces will also be discussed.; All the problems discussed above are very complex if the image is represented in a range-data format, as opposed to the gray-level format. This thesis contains a study of the associated functions, and details a complete functional prototype, the Manitoba Adaptor Registration System (MARS) that supports all these functionalities.