摘要
针对现有胸片中根据肺结节对病情诊断不精确的问题。采用一个基于拉普拉斯高斯滤波的参数化对数图像处理方法对CXR中的肺结点进行增强。该方法采用具有相应参数的Lo G来增强原始胸片中的结节状结构和边缘。然后再利用参数变化的PLIP方法提高图像对比度。文中选择熵值对此方法进行评估。熵值越小,表明图像增强的性能越好。从结果来看,采用不同参数的改进PLIP方法处理后图像的熵值平均为原始图像熵值的1/12。
Considered the disease diagnose by existing lung nodules in chest radiographs (CXRs) is not very accurate. In this paper, we proposed a parameterized logarithmic image processing (PLIP) method combined with the Laplacian of a Gaussian (LoG) filter to enhance lung nodules in CXRs. First we use LoG to sharp the edge of lung nodules and then apply the PLIP to highlight the contract. To evaluate the new method we introduced the measure- ment of enhancement by entropy evaluation to objectively assess this method. The smaller EMEE's value means the better image enhancement performance. When chose different c for improved PLIP, the average of image's EMEE is only 1/12 of original image's EMEE.
出处
《电子科技》
2017年第8期124-127,共4页
Electronic Science and Technology
关键词
胸片
肺结节
图像增强
参数化数图像处理
高斯的拉普拉斯
Chest Radiographs (CXRs)
lung nodules
image enhancement
Parameterized Logarithmic Image Processing (PLIP)
Laplace of Gaussian (LoG)