基于加权方差的煤矸石X射线图像分形维数最优估计算法

传统方法对于煤矸石X射线图像分形维数估计存在单层次缺陷,提出了基于加权方差的煤矸石X射线图像分形维数最优估计方法。由局域窗口特征检测方法获取图像的二维边缘像素特征分量,取其最大值构成图像采集的像素特征量,实现对图像的特征分解。在此基础上,提取二值化图像边缘轮廓特征量,采用包络轮廓线分割方法超分辨融合煤矸石X射线图像,建立煤矸石X射线图像的模板匹配模型。采用多层次纹理重建方法,获取图像的活动轮廓多层次分布集,结合加权方差获取图像方差和标准差,将图像的分形维数估计方差和标准差输入图像的相关性检测模板匹配函数中,进行图像分形维数最优估计。仿真结果表明,采用该方法进行煤矸石X射线图像分形维数估计的精度较高,自适应性较好,提高了煤矸石X射线图像的识别和检测能力。

Optimal Estimation Algorithm of Fractal Dimension of Coal Gangue X-ray Image Based on Weighted Variance

The traditional method has a single-level defect in the fractal dimension estimation of coal gangue X-ray image.The optimal estimation method of coal gangue X-ray image fractal dimension based on weighted variance is proposed.The two-dimensional edge pixel feature component of the image is obtained by the local window feature detection method,and the maximum value is taken to form the pixel feature quantity of the image acquisition,and the feature decomposition of the image is realized.On this basis,the feature quantity of the edge contour of the binarized image is extracted,and the X-ray image of the coal-fired stone X-ray is super-resolved by the envelope contour segmentation method,and a template matching model of the coal-fired X-ray image is established.The multi-level distribution set of the active outline of the image is obtained by using the multi-level texture reconstruction method,and the image variance and standard deviation are obtained by combining the weighted variance.The estimated variance of the fractal dimension of the image and the standard deviation are input into the correlation detection template matching function of the image,and the optimal estimation of the fractal dimension of the image is carried out.The simulation results show that the fractal dimension estimation of coal gangue X-ray image is accurate and adaptive,and the recognition and detection ability of coal gangue X-ray image is improved.