光学 精密工程, 2014, 22 (11): 3100, 网络出版: 2014-12-08   

微焦点X射线图像乘性加性混合噪声的去除

Denoising method of micro-focus X-ray images corrupted with mixed multiplicative and additive noises
高红霞 1,2,*吴丽璇 1,2徐寒 1,2康慧 3胡跃明 1,2
作者单位
1 华南理工大学 精密电子制造装备教育部工程研究中心, 广东 广州 510640
2 华南理工大学 自动化科学与工程学院, 广东 广州 510640
3 广东技术师范学院 自动化学院 广东 广州 510665
摘要
考虑微焦点X射线仪成像信噪比低, 混合噪声污染严重等问题, 提出了一种乘性、加性混合噪声去除方法。首先, 建立了含乘性、加性混合噪声的图像模型;其次, 基于总变分和稀疏表示原理分别构造了滤除加性噪声和乘性噪声的目标函数;最后, 应用显式差分算法和梯度投影算法分步滤除加性噪声和乘性噪声。实验结果显示, 与总变分去加性噪声方法相比, 该方法处理后的图像平滑区域均值与标准差比(MSR)平均提升了10.9%, 细节区域拉普拉斯梯度模(LS)平均提升了15.6%。这些结果表明: 本文算法不仅有效滤除了微焦点X射线图像的混合噪声, 并且较好地保留了图像细节特征, 能够满足集成电路内部缺陷检测对图像平滑度和细节清晰度的要求。
Abstract
In consideration of the lower imaging Signal to Noise Ratio(SNR) and serious mixed noise of a micro-focus X-ray inspector, a denoising method was proposed for the images corrupted with mixed multiplicative noise and additive noise. Firstly, an image model was established to represent the micro-focus X-ray images with mixed multiplicative and additive noises. Then, to remove the mixed noises, the objective functions were proposed based on the principle of total variation and sparse representation. Finally, the multiplicative noise and the additive noise were removed by explicit difference method and gradient projection in steps. Experiment results show that the proposed method enhances the Mean to Standard deviation Ratio(MSR) of the images by 10.9% in smooth areas, the Laplacian Sum(LS) by 15.6% in detail areas as compared with total variation algorithm for the additive noise model. The experiments demonstrate that the proposed method not only removes the mixed noises in X-ray image but also retains the details of the image edge. It meets the requirements of integrated circuit detection for image smoothness and detail definition.

高红霞, 吴丽璇, 徐寒, 康慧, 胡跃明. 微焦点X射线图像乘性加性混合噪声的去除[J]. 光学 精密工程, 2014, 22(11): 3100. GAO Hong-xia, WU Li-xuan, XU Han, KANG Hui, HU Yue-ming. Denoising method of micro-focus X-ray images corrupted with mixed multiplicative and additive noises[J]. Optics and Precision Engineering, 2014, 22(11): 3100.

本文已被 3 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!