On-line layer monitoring and defect inspection of model maker rapid prototyping system using vision technology

Rapid Prototyping (RP) technology, which is actually a 2D layer material processing technology, has been successfully employed to fabricate a 3D object, layer by layer. However, the process stability and reliability for most RP systems present an impressive problem, since the RP technology is relatively new for general industries. Hence, on-line inspection and monitoring to increase the reliability and stability of the RP process is an interesting research topic. Vision technology also has been successfully applied to on-line inspection and monitoring of industrial processes, but most applications are still limited to 2D patterns. Hence, vision technology has great potential to inspect and monitor the RP process, because both of these two technologies are characterized as 2D processes. A monochrome CCD camera and several image-processing algorithms, including filtering, smoothing, image difference, and texture discrimination algorithms, were employed in this research to capture the object image and identify defects. The detailed algorithm for the adaptive texture analysis in defect inspection was developed and presented. Also, the effects of image differences and adaptive texture analysis on defect inspection were evaluated. Defect inspection using adaptive texture analysis is recommended in this research, though the algorithm and computation are more complicated. Once a defect was identified, a compensated program was generated and transferred back on-line to fix the defect, layer by layer. Hence, on-line monitoring was achieved. The research results show promise for on-line layer monitoring and defect inspection of Model Maker RP system using vision technology.