论文部分内容阅读
采用激光照射粗糙工件表面形成散斑图像,通过自相关函数分析散斑图像的二阶统计特性,使用分形算法提取散斑图像自相关函数矩阵的参数,建立分形维数与表面粗糙度对应的样本集合。用最小二乘法多项式拟合该样本集合,得到散斑图像与表面粗糙度值间的多项式关系。实验结果表明,基于激光散斑分形维数的表面粗糙度测量方法是可行的且适用于在线高精度粗糙度检测,在检测时间上从数十秒级提高到秒级,检测精度达到微米级。
Speckle images were formed by irradiating rough surface with laser, the second-order statistical properties of speckle images were analyzed by autocorrelation function, and parameters of autocorrelation function matrix of speckle images were extracted by fractal algorithm. Samples corresponding to fractal dimension and surface roughness were established set. The least square polynomial was fitted to the sample set to obtain the polynomial relationship between the speckle image and the surface roughness. The experimental results show that the method of surface roughness measurement based on the laser speckle fractal dimension is feasible and suitable for on-line high-precision roughness detection. The detection time is improved from the tens of seconds to the second level, and the detection accuracy reaches the micron level.