电子元件焊接坏点红外图像检测仿真研究  

Simulation and esearch on the Infrared Image Detection of Electronic Components Welding bad Point

作　　者：王付军[1] 

机构地区：[1]聊城大学东昌学院,山东聊城252000

出　　处：《计算机仿真》2016年第6期429-432,共4页Computer Simulation

基　　金：山东省高等学校科技计划项目(J14LN87)

摘　　要：由于电子元件焊接坏点红外图像与正常焊点特征不同,焊接坏点特征中与外部信息的特征很容易发生混淆,造成坏点中的图像特征很难与所在背景分离,有此产生相似性干扰。传统的计算机视觉检测方法,在滤除图像噪声时受到干扰影响会使坏点图像边缘弱化,不能有效识别坏点特征。为提高焊接坏点检测的准确性,提出采用红外图像分析的电子元件焊接坏点检测方法,首先对电子元件红外图像利用改进区域生长法进行图像分割,然后采用中值滤波法消除焊点图像的随机噪声并引用拉普拉斯算子对焊点图像边缘部分进行锐化处理,最后对电子元件的焊点进行模式识别以完成焊接坏点的检测。仿真结果表明,改进方法能够有效检测出电子元件的焊接坏点,检测效率有了显著提高。A detection method for electronic components welding bad point was proposed based on infrared image analysis. First of all, the infrared image of electronic components was segmented using the improved region growing method Then a median filtering method was used to eliminate the random noise of welding point image, and the La- place operator was introduced to make sharpening processing for the edge of welding point image. Finally, the pattern of welding points of electronic components was recognized to complete the welding bad point detection. Simulation ex- perimental results show that the improved method can effectively detect the welding bad point of electronic compo- nents, and the testing efficiency has significant improvement.

关 键 词：红外图像 焊接坏点 区域生长法 中值滤波法 拉普拉斯算子 

分 类 号：TP332[自动化与计算机技术—计算机系统结构]