Title :
CAD-based inspection of 3D objects using range images
Author :
Newman, Timothy S. ; Jain, Anil K.
Author_Institution :
Dept. of Comput. Sci., Michigan State Univ., East Lansing, MI, USA
Abstract :
Presents a CAD-based inspection system for defect detection in range images of castings. The inspection algorithms are applicable to a class of objects, specifically to objects that contain planar or quadric surfaces, rather than being specialized to a single object. CAD model information is used in several stages of the system, including surface classification and inspection. Techniques for detection of common casting defects and for inspection of manufacturing features and dimensional tolerances are presented. The methods are applied to a large number of real range images of castings with real defects. The inspection system is able to correctly classify over 90% of the casting images
Keywords :
CAD; automatic optical inspection; casting; computer vision; image sequences; 3D objects; CAD-based inspection; casting defects; castings; defect detection; dimensional tolerances; inspection algorithms; manufacturing features; planar; quadric surfaces; range images; surface classification; Casting; Computer science; Design automation; Humans; Inspection; Machine vision; Manufacturing; Packaging; Quality control; Surface finishing;
Conference_Titel :
CAD-Based Vision Workshop, 1994., Proceedings of the 1994 Second
Conference_Location :
Champion, PA
Print_ISBN :
0-8186-5310-8
DOI :
10.1109/CADVIS.1994.284496
