岩石节理裂隙图像由于噪声多、色彩复杂而分割困难,同时岩石节理裂隙检测在工程应用中又具有重要意义。将彩色图像从RGB转换到CIELAB色彩空间后,计算其方差图像。并将原始图像与方差图像相减得到图像的势能向量,同时利用小波系数进行加权,以此抑制岩体纹理噪声。最后,通过图像形态学方法获得边缘图像。实验结果表明,算法可以较好地抑制纹理噪声并能精确定位,以及获得理想的岩石节理裂隙边缘图像。
Rock fracture edge detection is important in rock engineering.The color images are transformed from RGB space to CIELAB space,and then variance images are obtained.After calculating the force that is weighted by wavelet coefficient,the texture can be suppressed.The color edge image is obtained by a morphological gradient approach.Experimental results show the proposed method is efficient and robust for rock fracture edge detection.