轴对称件DR检测过程中趋势项的消除方法
The Method for Eliminating Trend of Axisymmetric Body in DR Testing
-
摘要: 数字射线成像系统的轴对称件检测过程中, 由于多种原因导致的探测器中心图像亮度大于边缘亮度, 产生了某种趋势项, 而需识别的缺陷又叠加在趋势项上, 增加了缺陷检测的难度。研究了一种基于经验模式分解的信号趋势项提取方法。通过理论分析与试验验证, 证明了与传统方法(如最小二乘法)相比, 该方法能够更准确地提取图像中存在的趋势信息, 提高缺陷的识别率。Abstract: In the process of testing axisymmetric workpiece based on the digital radiographic system, caused by several factors, the center image is more bright than the margins. Some trend is produced. However, the flaw need to be distinguished superimposes on the trend, which increase the flaw testing difficulty. A extracting method for trend of signal based on empirical mode decomposition(EMD) was investigated. Through the analyse and experiment, it proved that the method could extract the trend of images more precisely and enhance the recognition rate of flaw comparing with the traditional trend analysis methods(such as least square approximation).