City Research Online

Abstract

A number of signal and image processing methods have been developed and evaluated for applications in ultrasonic quantitative non-destructive testing of highly scattering materials. The work falls into three main areas: signal and image enhancement and signal detection.

Analyses have been made of the differential features of grain and defect scattering in the frequency, time and space domains. The insight gained has led to the development of novel multi-channel adaptive filtering approaches to enhance true defect signals and images. The methods used include: normalised least-mean square error adaptive filtering, minimum-variance distortionless response array-processing and two-dimensional adaptive Wiener filtering. Automatic detection of the enhanced signals is achieved using constant false alarm rate detectors adapted and developed from well-established radar techniques. Two approaches have been used: cell-averaging detection and automatically censored mean level detection.

The performances of all the approaches are evaluated by processing extensive sets of A-scan data from test blocks containing artificial targets and a real flaw. Comparisons between the new adaptive filtering and detection approaches and existing methods such as split spectrum processing and spatial averaging have been presented in a number of tables summarising performance over a set of 64 sequential A-scans. These results show that the new approaches can detect all the test targets, with near zero false alarms, representing a considerable improvement over the performance of existing methods. The results of array and image processing are presented as false colour B-scan images in which the visibility of defect images corrupted by grain scattering has been considerably enhanced. An important feature of the current work is that fixed processing parameters have been used throughout.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > TK Electrical engineering. Electronics Nuclear engineering
Departments:	School of Science & Technology > Engineering > Electrical & Electronic Engineering School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads