高密度电子电路频谱图的边缘检测算法

在利用电磁扫描技术检测高密度电路板的组装质量时,为了识别频谱图中的异常频率点,需要检测和提取频谱图中频谱曲线.而频谱图中频谱曲线的检测和提取效果直接影响异常频率点检测的准确度.本文致力于电路板频谱图的边缘检测方法的研究,提出了多窗口Canny自适应阈值边缘检测算法,即通过对频谱图进行窗口划分,在窗口内采用改进的梯度幅值的计算方法和自适应阈值Canny检测算法,实现了对频谱图的边缘检测.理论分析和实验结果表明,本文所提出的算法准确度高,可以得到较好的边缘检测结果,为识别频谱图中的异常频率点提供了更精准的频谱边缘图像.

An Algorithm of Edge Detection of the High-density Board Spectrum

In detecting the assembly quality of high-density circuit boards by electromagnetic scanning,the detection and extraction of the spectrum curve is necessary for the purpose of recognizing the aberrant frequency points on its spectrum curve.The result directly affects the accuracy of the detection of aberrant frequency points on the curve.The paper deals with the methods for detecting the edges on the frequency spectrogram of a circuit board,and proposed an algorithm named multi-window adaptive threshold Canny edge detection algorithm,in which by dividing the frequency spectrogram into smaller windows in which the improved method of calculating the gradient magnitude and the method of adaptive-threshold Canny edge are adopted so that the detection of edges is accomplished.Theoretical analysis and experimental results show that the proposed algorithm is high in accuracy,and gets a better result in edge detection,which provides more accurate frequency spectrograms for recognizing the aberrant frequency points.