论文部分内容阅读
目前,X射线检测技术广泛应用于元器件的缺陷检测。但X光图像低灰度、低对比度的特点以及气泡形状的任意性,使得X光图像检测提取气泡并确定其位置成为元器件检测中的一个难题。元器件气泡的检测质量直接影响到元器件检测的速度以及准确性。通过对现有图像问题的分析,提出一种快速提取气泡的方法。该方法首先对X射线图像进行高斯滤波、对比度拉伸等预处理,然后采用Canny边缘检测确定气泡边缘,再应用数学形态学膨胀、区域填充等图像处理方法最终得到完整的气泡轮廓。实验结果表明,此方法能很好地满足元器件检测的要求。
At present, X-ray detection technology is widely used in the detection of defects in components. However, the characteristics of low-intensity and low-contrast of X-ray images and the arbitrary shape of bubble make X-ray image detection and extraction of bubbles and determination of their positions become a challenge in component inspection. Component bubble detection quality directly affects the speed of component detection and accuracy. Through the analysis of the existing image problems, a rapid extraction method of bubbles is proposed. Firstly, the X-ray images were preprocessed by Gaussian filtering, contrast stretching and so on. Then the edge of the bubble was determined by Canny edge detection. Finally, a complete bubble contour was obtained by image processing methods such as mathematical morphological expansion and area filling. Experimental results show that this method can well meet the requirements of component testing.