摘要：Full-matrix capture (FMC) and total focusing method (TFM) based on ultrasonic phased array technology serve as important techniques for structural defect detection. However, the presence of artifacts often impedes the automatic acquisition of accurate, high-precision dimensional information from defect images. In this study, an ultrasonic phased array system was constructed, along with a supporting software platform for FMC and TFM imaging. A defect sizing method based on image segmentation and noise reduction is proposed. The integrated hardware-software system acquires TFM images containing artifacts and applies an image segmentation approach to remove artifacts in the edge regions. Subsequently, artifacts in the core region are eliminated using an adaptive image denoising method. The area of the defect region is then quantified based on the information derived from image pixels. Additionally, a contrast model based on image morphology is introduced to achieve size quantification of an approximate 2D model. Experimental validation was performed using artificially fabricated blind hole defects, yielding an average area quantification error of 8.57%. The elliptical area approximation achieved 87.50% consistency with the safe area, and the average error across all samples was 7.30%.

关键字：FMC-TFM; Noise reduction; Defect characteristics; Size quantification

ISSN号：0263-2241

卷、期、页：卷: 270

发表日期：2026-04-21

期刊分区（SCI为中科院分区）：二区

收录情况：SCI（科学引文索引印刷版）,EI（工程索引）,SCIE（科学引文索引网络版）

发表期刊名称：Measurement

参与作者：程经纬

通讯作者：李勇,王禹钦

第一作者：董绍华,魏昊天

论文类型：期刊论文

论文概要：李勇,董绍华,魏昊天,王禹钦,程经纬，Size quantification method for blind holes based on ultrasonic total focusing imaging method，Measurement，2026，卷: 270

论文题目：Size quantification method for blind holes based on ultrasonic total focusing imaging method