Three-dimensional solid object laser scanner

The aim of this thesis is to investigate the different components of a laser scanner for industrial applications, like quality control or future domestic applications such as the front end of a video phone system. The scanner implemented in this project is based on the line laser range finding principle. The laser beam of an ImW Helium-Neon laser is diffracted in a vertical line, which is projected on the object. The profile highlighted on the object surface is recorded from a Vidicon camera, digitised and stored in the computer. The object is rotated from a stepper motor and in this way a series of profiles is acquired. The image of the profile is processed in order to eliminate the noise and define it, as clearly as possible. Different calibration methods have been investigated in order to extract the three dimensional co-ordinates of the profile points. The system has a resolution of 0.7mm/pixel and an average error of 0.3%. Each profile curve is segmented into a series of straight line segments in order to reduce the amount of data stored. The Marching Cubes Algorithm has been implemented and enhanced to speed up the formation and connectivity of the triangular facets. The algorithm yields sub pixel accuracy representations of the object but it is slow and requires a lot of temporary storage for the triangle connections. Both of the previously mentioned algorithms result in a wireframe mesh representation of the object. An algorithm has been designed to measure the internal volume of the object from its wireframe representation with an average error of 0.8%. The algorithm is based on the principle of a parallelepiped scanning the object in a raster fashion. The total volume of the, object is the sum of the elementary parallelepiped volumes. The scanning speed of the algorithm depends on the slantness of the object facets. Finally, the reconstructed object is displayed on the screen with a CAD style software. The project has exposed the inadequacies of the PC DOS environment as a development system for machine vision research in terms of its storage capabilities and its video graphics subsystem. The implemented system compares favourably with similar commercial systems. The project has proved that it is possible to implement a laser scanner on a low cost budget with reasonable accuracy and resolution