Finding position and characterization of the defects in 3D point cloud

The inspection process has been an inseparable part of manufacturing to measure dimensions such as diameter, flatness, roundness and straightness of the parts. Besides, on some machined surfaces, it is required to measure roughness and identify surface defects. For defect detection, companies are still relying on visual inspection, which is very slow and labor-intensive. To overcome all challenges, interferometry instruments are used to acquire 3D images. Still, once a surface is acquired, the position and size of defects have to be found, and sometimes defect has to be classified. Novacam Technologies Inc. designs and manufactures fiber-based low coherence interferometry instruments capable of acquiring volumetric data in hard to reach spaces. The goal of the proposed project is to identify defects on the acquired point cloud file of the curved surfaces and find its position, orientation and size. The major and minor axis of each defect will be computed and profiles through the deepest point extracted. All the parameters related to the defect's position and characterization will be listed in a log file.

Sevin Samadi
Faculty Supervisor: 
Sivakuamr Narayanswamy;Javada Dargahi
Partner University: