将霍夫变换与最小二乘法相结合,研究对实验数据和图像处理中的二值边缘图进行直线拟合的方法。
首先,用霍夫变换剔除数据点集中的干扰点或噪声,并将分布在不同直线附近的点分离出来 然后,用最小二乘法
拟合各直线。该方法既解决了直接使用最小二乘法拟合时,拟合直线易受干扰点或噪声的影响和数据点分布在
多条直线附近而无法拟合的两个问题 同时也解决了直接使用霍夫变换时,拟合直线精度不高和直线段有效区间
不容易控制的问题。,The Hough transform and least squares method, to study the experimental data and image processing in the binary edge map method for fitting a straight line. First of all, the use of Hough transform concentration of data points removed or noise interference, and a straight line in the vicinity of different points separated　and then, using the straight-line least square method. This method not only solved directly when using the least square method, fitting a straight line susceptible to interference or noise and data points are located in the vicinity of more than a straight line and the two issues can not be fitted　while also addressing the direct use of Huo Cardiff transform, the accuracy of fitting straight line segments is not high and the effective range is not easy to control.