Residual-Based Surface Segmentation for Monitoring Topographic Variations

Research output: Contribution to journalArticlepeer-review

Abstract

Monitoring topographic variations in the engineered surface is crucial for quality engineers since the change in the surface finish is closely related to the performance of products. However, several challenging issues such as the existence of spatial autocorrelation within the surface, and changes of topographic features such as position and shape of peaks and valleys across defect-free surfaces make it difficult to monitor variations in the surface. In addition, existing monitoring approaches fail to detect local changes in the surface. In this article, we present a new approach for monitoring topographic variations in surfaces. We develop a residual-based separation deviation (RBSD) model to effectively identify local surface changes. Residuals are obtained through the fit surface prediction model, which characterizes the generic behavior of defect-free surfaces, and binarized by the RBSD model to distinguish the defective region where residuals are autocorrelated. A spatial randomness-based monitoring statistic is introduced to evaluate binary patterns in order to detect surface anomalies. Numerical simulation and a case study of the coated paper surface monitoring are provided to demonstrate the effectiveness of the proposed approach.

Original languageEnglish (US)
JournalIEEE Transactions on Automation Science and Engineering
DOIs
StateAccepted/In press - 2020

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Keywords

  • Anomaly detection
  • Data models
  • Monitoring
  • Rough surfaces
  • spatial randomness test
  • surface monitoring
  • surface quality
  • Surface roughness
  • Surface topography
  • surface topography variation.
  • Surface treatment

Fingerprint

Dive into the research topics of 'Residual-Based Surface Segmentation for Monitoring Topographic Variations'. Together they form a unique fingerprint.

Cite this