首页|基于改进LSD的HTCC板错位检测算法研究及应用

基于改进LSD的HTCC板错位检测算法研究及应用

扫码查看
为解决摄像头模组所用HTCC板脱离固定槽位所导致的HTCC板破裂以及后续零件安装不良等问题,提出一种基于改进LSD的HTCC板错位检测算法.首先应用伽马变换消除产品图像上的阴影;然后对图像进行双边滤波,与原始LSD算法的滤波相比,双边滤波可以保留更完整的边缘信息;提出改进映射关系且应用双线性插值的尺度缩放算法,消除图像的量化伪影,减少图像的失真;提出断线再连接算法,解决原始LSD算法存在的线段过分割问题;最后基于改进的LSD算法设计HTCC板错位检测算法,求得产品中错位的HTCC板的数量和位置.在工厂现场验证得出该方法的检测平均耗时在800 ms以内,检测准确率在97%以上,与原始LSD算法相比,检测耗时至少减少了 38%,准确率至少提高了 21%,该方法可以满足生产的实时性和准确性要求.
Research and application of HTCC board dislocation detection algorithm based on improved LSD
In order to solve the problems that HTCC board used in camera module breaks away from the fixed slot,and the subse-quent parts are poorly installed,a dislocation detection algorithm of HTCC board based on improved LSD was proposed.Firstly,gamma transform was used to eliminate the shadow on the product image.Then,bilateral filtering was applied to the image.Com-pared with LSD algorithm,bilateral filtering can keep more complete edge information.Then,a scaling algorithm based on bilinear interpolation and improved mapping relation was proposed to eliminate quantization artifacts and reduce image distortion.Then,a disconnection and reconnection algorithm was proposed to solve the problem of over-segmentation of line segments in LSD algo-rithm.Finally,based on the improved LSD algorithm,the dislocation detection algorithm of HTCC board was designed,and the number and position of dislocation HTCC board in the product were obtained.The factory verification shows that the average de-tection time of this method is less than 800 ms,and the detection accuracy is over 97%.Compared with the traditional LSD algo-rithm,the detection time is reduced by at least 38%,and the accuracy is improved by at least 21%.This method can meet the re-al-time and accuracy of production.

LSD algorithmbilateral filteringvisual inspectionHigh-Temperature Cofired Ceramic(HTCC)

尚会超、韩鑫磊、嵇长委、彭向前

展开 >

中原工学院机电学院,郑州 450007

湖南科技大学机电工程学院,湘潭 411201

LSD算法 双边滤波 视觉检测 高温共烧多层陶瓷

2022年度河南省高等学校重点科研项目资助项目

22A580002

2024

现代制造工程
北京机械工程学会 北京市机械工业局技术开发研究所

现代制造工程

CSTPCD北大核心
影响因子:0.374
ISSN:1671-3133
年,卷(期):2024.(1)
  • 6