|
Image Processing
|
|
Image Denoising |
|
Images are often corrupted by noise in acquisition and
transmission,which usually degrades the quality of images. However, various
image-related applications, such as aerospace, medical image analysis,
object detection etc., generally require effective noise suppression to
produce reliable results. Furthermore, denoising is often necessary as a
pre-processing for other image/vision tasks, e.g. compression, segmentation
and recognition. Therefore, denoising has been one of the most important and
widely studied problems in image processing and computer vision.The
objective of denoising is to remove the noise effectively while preserving
the original image details as much as possible. So far, many approaches have
been proposed to get rid of noise. We propose a non-local algorithm for image denoising in wavelet domain, in which the parameter of the noise is unknown. By computing the noise variance on the wavelet domain for the user specified sample area , we obtain the feature vector about different wavelet bands of the noise. Coupled with this feature vector and based on the properties of local performance both in time and spatial domain of wavelet, we bring forward a non-local algorithm in wavelet domain. For each wavelet coefficient in every subband, the algorithm first measures its similarity with other wavelet coefficients in the same subband. It then takes this similarity as the weight to modify the wavelet coefficients. Experiments demonstrate that our method can reduce the noise effectively without blurring the features of the image. For the non-local denoising approach presented by
Buades et al., remarkable denoising results are obtained at high expense of
computational cost. In this paper, a new algorithm that reduces the
computational cost for calculating the similarity of neighborhood windows is
proposed. We first introduce an approximate measure about the similarity of
neighborhood windows, then we use an efficient \emph{Summed Square Image} (SSI)
scheme and Fast Fourier |
| Examples: |
|
|
|
|
|
Our denosing result is comparable with CVPR 05's method, yet is about fifty times faster than it. From left to right: original Lena; noised Lena: denoising result by CVPR'05 method; result obtained with our mehtod. |
|
|
|
Another two denoising results generated by our method |
|
Paper: Yanli Liu, Jin Wang, Xi Chen, Yanwen Guo, Qunsheng Peng, A robust and fast Non-local Algorithm for Image Denoising. Journal of Computer Science and Technology, 2008, 23(2): 270-279 (JCST, SCI, EI). Jin Wang, Yanwen Guo, Yiting Ying, Yanli Liu, Qunsheng Peng, Fast Non-Local Algorithm for Image Denoising. To Appear in IEEE International Conference on Image Processing 2006 (ICIP 2006), Atlanta, USA. To appear. (EI) Non-Local Image Denoising in Wavelet Domain (小波域中的非局部图像去噪方法). 提交至<电子学报>. |
|
|
|
Tonal Value Adjustment |
| Examples: |
|
|
|
|
|
Paper: In preparation. |
|
|
| Image/Video Completion |
| Examples: |
|
|
|
The guy in the top images is Yiting Ying, who is now in National Instrument Shanghai (美国国家仪器公司) and won the senond prize in Chinese "The 10th Challenge Cup" competition. (挑战杯全国二等奖, 浙江省特等奖) |
| Paper: Chunxiao Liu, Yingzhen Yang, Qunsheng Peng, Yanwen Guo. A New Distortion Minimization Approach for Image Completion based on a Large Displacement View, Computer Graphics International 2008. Chunxiao Liu, Yanwen Guo, Liang Pan, Qunsheng Peng, Fuyan Zhang, Image Completion based on Views of Large Displacement. The Visual Computer, 2007, 23(10): 833-841. (TVC, SCI, EI). Chunxiao Liu, Liang Pan, Yanwen Guo, Jin Wang, Weichen, Qunsheng Peng, Image Inpainting Based on Large Displacement View Images. Journal of Software, 2006, 17:138-147. (In Chinese) To appear (刘春晓, 潘梁, 郭延文, 王进, 彭群生, 基于大位移视点图像的单帧图像修复技术. 软件学报, 17:138-147.) (EI) Another cooperative paper with Dr. Chunxiao Liu about image completion is in preparation. |
| User study of image retargeting(Link)
|
| Back to the homepage |
