Time filter

Source Type

Hua Z.,Shandong Institute of Economic and Techonlogy | Li Y.,Shandong Institute of Economic and Techonlogy | Li J.,China Institute of Technology
2010 6th International Conference on Wireless Communications, Networking and Mobile Computing, WiCOM 2010 | Year: 2010

The essence of image segmentation is a based on some properties the process for pixel classification. Firstly, three typical methods in image segmentation methods are outlined and their characteristic is analyzed in this paper. The traditional visual attention model is described and improved in this paper. The input image gray value and edge features are extracted by Gabor filters and the Gauss - Laplace operator, gray feature maps and the edge feature maps are got respectively, then interest regions image is obtained by linear combination. The interest region in interest region image is selected by the dynamic neural network methods in artificial intelligence. The limit scope of regional growth is provided improved visual attention model algorithm identified interest region, the binary image is got by setting gray-value. At last, image segmentation is achieved by image segmentation algorithm based improved visual attention model and region growing. Experimental results validate that this methods not only achieve image segmentation, but also accurately and automatically achieve interest region segmentation, improve the quality of the segmentation, and has good robustness. © 2010 IEEE.


Hua Z.,Shandong Institute of Economic and Techonlogy | Li Y.,Shandong Normal University | Li J.,China Institute of Technology
International Journal of Digital Content Technology and its Applications | Year: 2011

The contourlet transform possess flexible multi-scale and multi-directional property, and nonsubsampled contourlet transform (NSCT) possess the shift-invariant, which conducive to image enhancement research. Therefore, an image nonlinear enhancement algorithm based on nonsubsampled contourlet transform is proposed. The new algorithm has well completed the nonsubsampled contourlet transform, and respectively realized the image enhancement of the lowfrequency sub-band and high-frequency sub-band. In order to enhance image details information, the edge scale of the high-frequency sub-band is enhanced. In order to improve the quality of image enhancement, the intensity, hue and saturation of the low-frequency sub-band are respectively enhanced. At last, the enhanced coefficients will be nonsubsampled contourlet inverse transformed to obtain the enhanced image. The experimental results show that this paper algorithm can improve the structural property and the edge information of the image and improve the quality of image enhancement.


Hua Z.,Shandong Institute of Economic and Techonlogy | Li Y.,Shandong Institute of Economic and Techonlogy | Li J.,China Institute of Technology
2010 International Conference on Computer Design and Applications, ICCDA 2010 | Year: 2010

Image inpainting is an important content in the image recovery research area, and also is a research hot spot in the computer graphics and computer vision in the current. Two major categories image inpainting techniques are introduced. After Pre-processing, we get gray image and using contour extraction algorithm to extract the contour features of damaged area, which determine the contour feature location. Area-based matching algorithm is used to search alternative matching area. Contour-based Similarity Distance function (CSD) is used to determine the most similar matching area. According to the damaged area and the matching area, the source cloning domains and target cloning domains of MVSC algorithm is improved. Using mean value coordinates to achieve regional pixel cloning from the source cloning domain to the target cloning domain. Finally, the improved MVSC algorithm is used to repair the image. Experimental results verified that the method can not only repair small-scale scratches, but also repair a large area damaged areas. © 2010 IEEE.


Li J.,Shandong Institute of Economic and Techonlogy | Li Y.,Shandong Normal University | An Z.,Shandong Institute of Economic and Techonlogy
Journal of Convergence Information Technology | Year: 2011

In order to improve image denoising effect, an image denoising algorithm based on the nonsampled double density Contourlet transform was proposed. This algorithm firstly use the nonsampled double density Contourlet transform (NSDDCT) to decompose the input image, so as to obtain the highfrequency sub-band and the low-frequency sub-band. For high-frequency sub-band coefficient, this paper adopts gaussian denoising rules to raise the image definition; For low-frequency sub-band coefficient, this paper adopt threshold denoising rules to improve the visual effect. At last, the denoised image is obtained after reconstructing the decomposed sub-band. The experimental results and the objective evaluation results show that this method can not only effectively eliminate the noise of the image, but also can improve image denoising effect, obtain more clear images.


Hua Z.,Shandong Institute of Economic and Techonlogy | Li Y.,Shandong Institute of Economic and Techonlogy | Li J.,Shandong Normal University
International Journal of Advancements in Computing Technology | Year: 2011

Visual attention model has a good visual attention characteristic, which can accurately detect the salient region in images. Image scaling algorithm based on salient region will be studied. The salient region in original images is firstly detected, and the different region will be resized by the different scaling rules. At last, the resized image will be obtained by the fitting function. The salient region is resized by the nonlinear scaling rules adjust the scaling weights, which avoid the linear scaling rules causing morphing and ensuring the visual attention regional image not distortion. The other region is resized by the linear scaling rules use the interpolation method. Experimental results validate that this paper algorithm has not only achieved image scaling, and ensure the image no distortion and the integrity of salient region content, show good visual effect, and improve the quality of image scaling.


Hua Z.,Shandong Institute of Economic and Techonlogy | Li Y.,Shandong Normal University | Li J.,Shandong Normal University
Proceedings - 2010 7th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2010 | Year: 2010

Based on scale-invariant feature transform (SIFT) and mean seamless cloning (MVSC) , an image stitching algorithm is presented, to improve the quality of the panoramic stiching image. Using SIFT algorithm to extract between benchmark images (await matched image) and follow-up images(with the baseline image match the image) of the feature points, identifying locations and directions, using 128 dimensional vector to describe features point. Using the nearest neighbor method to achieve two images feature point matching, identify overlap regions. Using SIFT algorithm to provide benchmark images and follow-up images to determine the source cloning domain and target cloning domain of the MVSC. Using the mean value coordinates to achieve the pixel to interpolat from the source cloning domain to target cloning domain. Finally, using MVSC algorithm to achieve the two images stitching. Experiment results shows that this method with regard to image rotation, perspective changes and image scaling to have a very good stitching results, stitching image is complete information, the quality of the image is high. ©2010 IEEE.

Loading Shandong Institute of Economic and Techonlogy collaborators
Loading Shandong Institute of Economic and Techonlogy collaborators