17–22 May 2026
C.I.D
Europe/Zurich timezone

Development of a Computer Vision Based Inner Surface Inspection System for Cavity

TUP7631
19 May 2026, 16:00
2h
C.I.D

C.I.D

Deauville, France
Poster Presentation MC7.T07: Superconducting RF Poster session

Speaker

SeongJae Choi (Korea University)

Description

A cavity inner surface inspection system integrated with computer vision was developed to detect manufacturing defects such as pit, dent, and welding spatter that can degrade the RF performance of superconducting cavities. The system consists of high precision linear actuators, a high resolution CCD camera, an optical 45-degree mirror, and an addressable LED ring light. With individually controlled 12 LEDs on ring light, the system acquires a set of different illumination images at each region of interest. From these images, surface information including coating uniformity, roughness, grain boundaries, and three-dimensional defect morphology is extracted through analysis of albedo and normal vector maps based on photometric stereo. Accurate two-dimensional defect width and length are obtained by converting image coordinates into world coordinates using the x- and y-direction actuators. Defect height and depth are reconstructed using a depth-from-focus approach implemented with the z-direction actuator. The two-dimensional width and length resolution was 11.6 ㎛ per pixel.
This inspection system enables quantitative classification of surface defects and applied throughout the cavity manufacturing process to improve quality control and reduce performance-limiting surface irregularities.

In which format do you inted to submit your paper? Word

Author

SeongJae Choi (Korea University)

Co-authors

Prof. Eun-San Kim (Korea University Sejong Campus) Jongmo Hwang (Korea University Sejong Campus)

Presentation materials

There are no materials yet.