报告题目：Thermographic Data Processing for Material Defect Detection

报告时间：2019年6月26日（周三）下午14:00

报告地点：C1-435

主讲人：姚远 博士、副教授

主讲人简介：

摘要：Infrared thermography (IRT), widely used as a nondestructive testing (NDT) method, offers many advantages for material defect detection, such as quick inspection, low cost, wide scanning range and easy operation. The principle of IRT is as follows. First, an external energy source is used to heat the surface of the tested specimen. The heat flux then diffused into the inside of the material. In the meantime, the surface temperature response of the specimen is captured using an infrared camera as a time series of thermal images. By investigating these thermal images, the locations of defects can be identified by quality engineers from the contrast between normal and defective areas. However, in practice, the factors including noise, external environmental disturbances, and uneven heating may lead to low level of contrasts in the collected thermal images. In attempts to deal with these issues, multiple types of thermographic data processing methods have been proposed, which can be roughly divided into three groups: noise reduction/background elimination, thermal image segmentation, and direct feature extraction. In this talk, the fundamentals of IRT will be briefly introduced and the typical thermographic data processing methods will be reviewed with some application results. It is expected that the statistical and machine learning methods will be more widely applied in the field of NDT.

主讲人简介：姚远博士，清华大学（台湾）化学工程学系副教授。于2001年和2004年获得浙江大学控制科学与工程学系的学士和硕士学位。其后赴香港科技大学化学工程与生物分子工程学系深造，于2009年取得博士学位。同年至香港科技大学高分子成型过程及系统中心任副研究员(Research Associate)从事博士后研究，并于2011年进入清华大学（台湾）化学工程学系任教，从事过程数据分析相关研究。2016年获得“科技部”（台湾）优秀年轻学者研究计划。主要研究领域为将统计方法及人工智能技术应用于化工生产过程的分析与改善，并结合过程控制，从而确保过程安全，提升生产效率及产品良率。具体研究方向包括但不限于：过程数据分析与监测、故障检测及诊断、软测量技术、复合材料与文化遗产之非破坏性检测、计算流体力学模拟、树脂传递模塑过程、控制器性能评估等。其研究成果均发表于所属研究领域之国际顶尖期刊。姚远副教授从事科研以来，截止2019年6月，共发表SCI论文73篇，专书章节2篇，及获得5项发明专利。

自动化与电气工程学院科研处

2019年6月24日