搅拌摩擦连接区域表面弧纹图像处理与间距特征提取

Processing and Characteristic Parameter Procurement of Surface Arc Grain Image of Joint Area of Friction Stir Joining

弧纹是搅拌摩擦连接区域表面的重要的特征形貌之一,而间距是弧纹的主要几何参数。本文使用图像识别的方法对弧纹间距进行检测,提出了针对弧纹特点的图像获取方法,并通过对此方法下获得的原始图像进行灰度化、二值化、二值图像后续优化等处理方法得到可识别的二值图像。对于所得到的二值图像进行特征区域截取,在此特征区域内进行弧纹间距参数的检测并最终得到所要检测的弧纹间距参数。本文所使用的方法具有较高的识别成功率且算法可靠,可计算性能好。

The arc grain is a main characteristic in the surface of the joint area of friction stir joining,and the spacing is an important parameter of arc grain. A method of measuring arc grain spacing which based on image recognition is proposed, and a method for taking arc grain image according to its appearance characteristic is put forward. Then through the processing of gray processing, binaryzation, binary image optimization, this original image can be converted into a processable binary image.And after extracted the processable area in the binary image, the spacing of arc grain can be calculated out by using the proposed algorithm. This arc grain spacing measuring method proposed in this paper has a high success identify rate and reliability, and the algorithm of this method is easy to implement.