Abstract: Aiming at problem which is not beneficial for a physician to diagnose lung nodules in a chest X-ray image，such as lower image contrast，poorer visual effect，a Laplacian of Gaussian Filter technique based on visual perception and directional scale for enhancing lung nodules in chest X-rays is proposed.The implementation of the proposed method is based on the modification of classical Laplacian of Gaussian（LoG） filter operation by assigning different scale values in different directions and consider visual factor to complete vision correction.Experiments show that the technique proposed in this paper can effectively enhance the lung nodules in the chest radiograph，and with respective to the methods of the literature，the proposed method have better enhanced effect.

Key words: chest X-rays, lung nodules, image enhancement, visual perception, directional scale, Laplacian of Gaussian（LoG） filter

摘要： 针对医学X胸片影像对比度低，视觉效果差，不利于临床医师对X胸片肺部节点进行检诊的问题，在对X胸片肺部节点图像特点以及传统拉普拉斯高斯图像增强算子深入分析的基础上，提出一种基于视觉及方向尺度拉普拉斯高斯（Laplace of Gaussion，LoG）算子的医学X胸片肺结节增强方法。该方法是在传统拉普拉斯高斯滤波算子的基础上，通过对不同方向纹理引入不同算子尺度，并考虑视觉因素，进行视觉矫正实现的。实验表明，该文算法能有效增强X胸片肺部节点影像。与已有方法相比，该方法具有更好的增强效果。

关键词: X胸片, 肺结节, 图像增强, 视觉, 方向尺度, 拉普拉斯高斯