摘要
针对印刷线路板线序识别故障检测抗干扰能力弱和实时性低等问题,提出了一种基于高斯混合模型的自适应线序识别检测算法。通过图像采集和图像预处理得到高质量图像;利用高斯金字塔模板匹配算法实现对颜色特征区域的快速定位,提取线束所有像素点的颜色特征;利用基于期望最大化算法的高斯混合模型的无监督学习算法进行线序颜色分类。实验结果表明,系统线束定位速度平均耗时9.7 ms,颜色分类实验耗时13.6 ms,符合工业检测的实时性要求。该系统可以实现线束快速自适应定位,线序颜色识别准确,检测抗外界干扰能力强。
To address the problems of weak anti-interference capability and low real-time performance in PCB board wire sequence identification and fault detection,an adaptive wire sequence identification and detection algorithm based on the Gaussian mixture model is proposed.High-quality images are obtained through image acquisition and image preprocessing;the Gaussian pyramid template matching algorithm is used to achieve rapid positioning of color feature regions and extract the color features of all pixel points of the wire harness;the unsupervised learning algorithm based on the Gaussian mixture model with the EM algorithm is applied for wire sequence color classification.The experimental results show that the average time consumed for the wire harness positioning of the system is 9.7 ms,and the time consumed for the color classification experiment is 13.6 ms,which meets the real-time requirements of industrial detection.The designed system can realize rapid and adaptive positioning of the wire harness,accurately identify the color of the wire sequence,and accurately distinguish faulty wire harnesses.
作者
李兴民
赵河明
晏永
LI Xingmin;ZHAO Heming;YAN Yong(College of Mechanical and Electrical Engineering,North University of China,Taiyuan 030051,China;Shanxi Key Laboratory of High-End Equipment Reliability Technology,Taiyuan 030051,China)
出处
《测试技术学报》
2025年第5期573-580,598,共9页
Journal of Test and Measurement Technology
基金
中北大学高端装备可靠性技术山西省重点实验室研究基金资助项目(446-110103)。
关键词
装备可靠性
线序检测
金字塔模板匹配
特征提取
高斯混合模型
equipment reliability
line order detection
pyramid template matching
feature extraction
Gaussian mixture model(GMM)
