| 张婷婷,马杰,刘新艳,葛岭岭.低秩遮挡图像去噪方法[J].光电子激光,2015,26(1):184~191 |
| 低秩遮挡图像去噪方法 |
| Denoising method for low rank occluded images |
| 投稿时间：2014-07-18 |
| DOI： |
| 中文关键词: 低秩矩阵 矩阵填充 矩阵恢复 非精确增广拉格朗日乘子(IALM) |
| 英文关键词:low-rank matrix matrix completion matrix recovery inexactly augmented Lagra nge multiplier (IALM) |
| 基金项目:国家自然科学基金(61203245)和河北省自然科学基金(F2012202027)资助项目 (河北工业大学 电子信息工程学院,天津 300401) |
|
| 摘要点击次数: 1430 |
| 全文下载次数: 70 |
| 中文摘要: |
| 为获得清晰的低秩图像,提出一种将低秩矩阵填充 (LRMC)与低秩矩阵恢 复(LRMR)联合的新模型,基于非精确增广拉格朗日乘子(IALM)法进行求解,运用LRMC去除遮 挡并填充缺失部 分,再利用LRMR去除噪声,得到完整的图像。以恢复时间、信噪比(SNR)、峰 值信噪比(PSNR)、差错率(err)等做评价标准,对3幅受噪声污染的图像的恢复结果表明, 本文提出的联合LRMC与LRMR的新模型,既能去除遮挡又能够填充图像的缺失部 分,能够达到理想的恢复效果。 |
| 英文摘要: |
| Both deletion and the noise pollution may exist in low-rank images, that is to say, both data loss and destruction occur in low-rank images.In view of the existing problems,this paper proposes a new model combining matrix recovery with matrix completion based on inexactly augmen ted Lagrange multiplier (IALM).It removes the cover and fills the missing parts using low-r ank matrix completion,and simultaneously it gets rid of the noise to obtain the full image using low-rank matrix recovery.For the three different images with noise,comparing their recovery tim e,signal-to-noise ratios,peak signal-to-noise ratios and error rates with those of the existing low-rank matrix algorithm, it can be seen that the improved low-rank matrix completion with recovery model can get rid of the cover noise images very well.In the repairing of obscured images,change the model and recover the cover images.The low-rank matrix completion based on low-rank matrix recovery can remove the cover and fill the missing parts simultaneously. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|