레이저 스캐너를 이용한 자동차 부품 품질 측정 시스템 개발 : Development of a Quality Inspection System for Automotive Parts Using Laser Based Scanner
툴라다르 우펜드라 마니
- 원문 URL
This dissertation presents a method to use laser scanned data for the inspection of complex surfaces of automotive parts. The method involves processing of laser scanned three-dimensional (3D) point cloud data and then comparison of data with respect to reference model. For the inspection of free form surface, a surface matching process is used whereas for the inspection of flat surfaces, a comparison with reference plane-surface is introduced. For the construction of reference plane surface, surface fitting method is adopted. Surface matching is the process that compares surfaces and decides whether they are similar. In 3D computer vision, surface matching plays a prominent role. Surface matching has a wide range of application such as object recognition, 3D model construction from patches of scanned surfaces, comparison between scanned model and CAD models and many more. By comparing two surfaces, an association between a known object and sensed data is established. By computing the 3D transformation that aligns two surfaces, surface matching can also be used for surface registration. Since the coordinate systems of two surfaces to be compared are undefined, the surface matching is a challenging task. Therefore the typical approach to surface matching is to transform the surfaces being compared into representations where comparison of surfaces is straightforward. Surface matching is further complicated by characteristics of sensed data, including clutter, occlusion and sensor noise. This research work approaches a modified iterative closest point (ICP) algorithm to solve the problem of surface matching. Surface reconstruction is widely used for constructing an optimal surface using point cloud data. It is also used to construct the missing surfaces from the parts by analyzing the geometric properties and information. This technology is mainly used in reverse engineering. In this research we deal with the planner approximation technique rather than interpolation technique. For the planar approximation the scanned point cloud data of the flat surface of work piece are used. A reference plane parallel to the direction of the scanned flat surface is defined at the mean position of the flat surface. The direction of the reference plane is determined by the normal vector of the flat surface. Construction of the reference plane is carried out by the singular value decomposition (SVD) technique. Then the deviation of the scanned surface from the reference plane is measured by calculating the error distance between the points of the surface to the reference plane using the least-squares method.