超声相控阵PE检测图像缺陷定位算法研究

Study on flaw orientation algorithm of ultrasonic phased array PE detection image

针对车轮无损检测图像中对缺陷评判的不足,提出一种能够自动定位检测图像中缺陷的算法。该算法是在相控阵技术的车轮超声探伤基础上,运用数字图像处理技术,对检测图像进行时域的处理。对算法的流程进行了详细介绍,并对每一步给出说明和处理结果。通过对比得知,这一算法对图像中缺陷的检测精度高于90%,大为提高识别效率,并减少人为因素引起的误判。

In order to overcome the shortcomings of flaw evaluations to nondestructive detecting images of wheels, an algorithm was developed to orientate the positions of flaws in images automatically. This algorithm is based on ultrasonic test of wheels with phased array technology, using digital image processing technology in the time - domain. This paper introduces the processes of the algorithm, and makes an instruction to each step and shows the result, comparison with the standard images shows that the accuracy of the positioning algorithm is higher than 90%. The algorithm greatly improves the recognition efficiency and reduces incorrect judgments caused by human factors.