基于FPGA的激光拼焊板焊缝位置快速检测方法

FPGA based Rapid detection method for Laser Weld Seam in Robotic Tailored Blank Welding

提出了基于microblaze软核处理器的SOPC技术来进行焊缝结构光图像处理与特征提取;由图像预处理模块实现对焊缝图像滤波、二值化、数学形态学处理、边缘检测、中心线提取等步骤的预处理,预处理后的图像送入MicroBlaze软核进行特征点提取。该方法可以准确而快速的提取图像焊缝位置。相比传统的处理方法,可以大大提高焊缝跟踪系统的实时性。

A SOPC based on soft-core MicroBlaze processor is applied to image process and feature extraction of laser welding seam. The image-processing IP core based on OPB bus in the SOPC is self-defined to implement algorithms such as filtering enhancement, image binarization, mathematic morphological operation, edge detection and centerline extraction. Then the data is transferred through OPB bus to MicroBlaze to do the feature extraction. The experiment indicates that this design can improve the real time ability of the weld seam tracking, and also hold the features of flexibility and reconfiguration.