基于LBF模型的斑点噪声图像分割

Speckle Noise Image Segmentation Based on LBF Model

LBF模型的核函数的尺度函数是一个固定的任意值,不能实现在不同区域采用不同的尺度。不同点的灰度均匀性不同,因此仅用一个固定尺度的模型来计算不同区域的统计信息是不准确的。针对这一问题论文提出了自适应的高斯核函数,在图像的不同区域,采用不同的尺度参数,使图像的能量差更加明显。实验结果表明,论文提出的方法相较于传统LBF模型和CV模型,提高了斑点噪声图像的分割精度。

The scale function of the kernel function of the LBF model is a fixed arbitrary value,and different scales can't be used in different regions.Different points have different gray uniformity,so it is inaccurate to use only one fixed-scale model to calculate the statistics of different areas.In response to this problem,an adaptive Gaussian kernel function is proposed.Using different scale parameters in different regions of the image makes the energy difference of the image more obvious.The experimental results show that compared with the traditional LBF model and CV model,the proposed method improves the accuracy of speckle noise image segmentation.