Time filter

Source Type

Pan X.-Y.,Xi'an University of Science and Technology | Pan X.-Y.,Xidian University | Pan X.-Y.,Key Laboratory of Intelligent Perception and Image Understanding | Liu F.,Xidian University | And 2 more authors.
Ruan Jian Xue Bao/Journal of Software | Year: 2010

By using the density sensitive distance as the similarity measurement, an algorithm of Density Sensitive based Multi-Agent Evolutionary Clustering (DSMAEC), based on multi-agent evolution, is proposed in this paper. DSMAEC designs a new connection based encoding, and the clustering results can be obtained by the process of decoding directly. It does not require the number of clusters to be known beforehand and overcomes the dependence of the domain knowledge. Aim at solving the clustering problem, three effective evolutionary operators are designed for competition, cooperation, and self-learning of an agent. Some experiments about artificial data, UCI data, and synthetic texture images are tested. These results show that DSMAEC can confirm the number of clusters automatically, tackle the data with different structures, and satisfy the diverse clustering request. © by Institute of Software, the Chinese Academy of Sciences.


Wang S.,Key Laboratory of Intelligent Perception and Image Understanding | Liu K.,Key Laboratory of Intelligent Perception and Image Understanding | Pei J.,Key Laboratory of Intelligent Perception and Image Understanding | Gong M.,Key Laboratory of Intelligent Perception and Image Understanding | Liu Y.,Key Laboratory of Intelligent Perception and Image Understanding
IEEE Geoscience and Remote Sensing Letters | Year: 2013

This letter presents a new unsupervised classification method for polarimetric synthetic aperture radar (POLSAR) images. Its novelties are reflected in three aspects: First, the scattering power entropy and the copolarized ratio are combined to produce initial segmentation. Second, an improved reduction technique is applied to the initial segmentation to obtain the desired number of categories. Finally, to improve the representation of each category, the data sets are classified by an iterative algorithm based on a complex Wishart density function. By using complementary information from the scattering power entropy and the copolarized ratio, the proposed method can increase the separability of terrains, which can be of benefit to POLSAR image processing. Three real POLSAR images, including the RADARSAT-2 C-band fully POLSAR image of western Xi'an, China, are used in the experiments. Compared with the other three state-of-the-art methods, H/\alpha$-Wishart method, Lee category-preserving classification method, and Freeman decomposition combined with the scattering entropy method, the final classification map based on the proposed method shows improvements in the accuracy and efficiency of the classification. Moreover, high adaptability and better connectivity are observed. © 2004-2012 IEEE.


Hou B.,Key Laboratory of Intelligent Perception and Image Understanding | Cheng X.,Key Laboratory of Intelligent Perception and Image Understanding | Jiang H.Q.,Key Laboratory of Intelligent Perception and Image Understanding
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2012

The compressive sensing theory shows that signals and images can be recovered from far fewer samples than that used in Shannon sampling theorem. In practical applications our aim is that an object should be meaningfully reconstructed in considerable detail relative to the full scene elsewhere at low sampling rate so that the target will be found directly from the reconstructed image. To achieve the goal, a new method based on manifold-based compressive sensing is proposed for object specific image reconstruction, in which the whole image is divided into small pieces and reconstructed piece by piece with the probability density function of the target as prior knowledge. Our reconstruction method is very fast since we divided the whole image into small pieces and the small pieces are reconstructed respectively. In our method, modeling a manifold of a target and getting the probability density function of the target is the key issue. And the model, which is used to obtain the probability density function about the target, is a Mixture of Factor Analyzers (MFA). The experiments results show that the target can be reconstructed more clearly than the full scene elsewhere at the low sampling rate with our method. © 2012 Springer-Verlag.


Wu J.,Key Laboratory of Intelligent Perception and Image Understanding | Qi F.,Xidian University | Shi G.,Xidian University
APSIPA ASC 2011 - Asia-Pacific Signal and Information Processing Association Annual Summit and Conference 2011 | Year: 2011

In this paper, we introduce a unified spatial masking function for the estimation of just-noticeable difference (JND). Conventional models estimate several parts independently, and then combine these parts to get the JND. In this work, we treat the spatial masking effect as a nonlinear transformation of the luminance adaptation. To model the transformation, we measure the deviation of image contents from the ideal patterns to establish luminance adaptation rules. Considering both luminance difference and structural regularity, we derive a nonlinear spatial masking function by modulating luminance adaptation with the deviation coefficients. The masking function deduces an accurate estimation of the JND. Experiments demonstrate the validity of the proposed framework.


Wu J.,China Jiliang University | Wu J.,Xidian University | Wu J.,Key Laboratory of Intelligent Perception and Image Understanding | Liu F.,Xidian University | And 2 more authors.
2012 5th International Congress on Image and Signal Processing, CISP 2012 | Year: 2012

The prior information of image plays an important role in compressive sensing (CS) reconstruction. The edge is one of the important information which exists in the image to be recovered. In this work, the edge of images is used as a prior information and extracted from CS measurements in the wavelet-based CS inversion. An edge-based matching pursuit algorithm is developed by merging edge information to guide the pursuit process. Edge information well defines the locations of the significant coefficients to be recovered, thus it contributes so much to the reconstruction quality. Experiment results demonstrate that the proposed algorithm returns superior reconstruction performance for the images with obvious edges and high sparsity compared with other CS algorithms. © 2012 IEEE.


Shi C.,Xidian University | Shi C.,Key Laboratory of Intelligent Perception and Image Understanding | Liu F.,Xidian University | Liu F.,Key Laboratory of Intelligent Perception and Image Understanding | And 8 more authors.
IEEE Transactions on Geoscience and Remote Sensing | Year: 2015

Although the bandwidth of the high-resolution panchromatic (HR PAN) image is wide, it is narrow in each band of the low-resolution multispectral (LR MS) image. Hence, the spatial resolution of the HR PAN image is much higher than that of the LR MS image. However, HR PAN image only has a single band. The purpose of the Pan-sharpening algorithm is to make the Pan-sharpened image with both high spatial resolution and good spectral information. In this paper, a novel learning interpolation method for Pan-sharpening is proposed by expanding the sketch information in the HR PAN image. The sketch information contains the edges and lines features of the image, and each segment of the sketch information has its own direction. According to the primal sketch graph of the HR PAN image, a regional map is obtained by a designed geometrical template. Since the size of the HR PAN image is different from that of the LR MS image, the LR MS image is interpolated into an interpolated multispectral (IMS) image by the nearest interpolation method. In addition, the IMS image can be mapped into the structure and the nonstructure regions by this regional map. The nonstructure regions are divided into the smooth and the texture regions by a variance value. For the structure and texture regions, the interpolated pixels in the IMS image are relearned and readjusted by the proposed structure and texture learning interpolation method, respectively. Experimental results show that the proposed Pan-sharpening method can provide superior performance in both visual effect and quality metrics, particularly for the images with a large spectral difference. © 1980-2012 IEEE.


Qi Y.,Xidian University | Qi Y.,Key Laboratory of Intelligent Perception and Image Understanding | Liu F.,Xidian University | Liu F.,Key Laboratory of Intelligent Perception and Image Understanding | And 4 more authors.
Applied Soft Computing Journal | Year: 2012

By replacing the selection component, a well researched evolutionary algorithm for scalar optimization problems (SOPs) can be directly used to solve multi-objective optimization problems (MOPs). Therefore, in most of existing multi-objective evolutionary algorithms (MOEAs), selection and diversity maintenance have attracted a lot of research effort. However, conventional reproduction operators designed for SOPs might not be suitable for MOPs due to the different optima structures between them. At present, few works have been done to improve the searching efficiency of MOEAs according to the characteristic of MOPs. Based on the regularity of continues MOPs, a Baldwinian learning strategy is designed for improving the nondominated neighbor immune algorithm and a multi-objective immune algorithm with Baldwinian learning (MIAB) is proposed in this study. The Baldwinian learning strategy extracts the evolving environment of current population by building a probability distribution model and generates a predictive improving direction by combining the environment information and the evolving history of the parent individual. Experimental results based on ten representative benchmark problems indicate that, MIAB outperforms the original immune algorithm, it performs better or similarly the other two outstanding approached NSGAII and MOEA/D in solution quality on most of the eight testing MOPs. The efficiency of the proposed Baldwinian learning strategy has also been experimentally investigated in this work. © 2012 Elsevier B.V.


Liu C.,Key Laboratory of Intelligent Perception and Image Understanding | Shen C.,Xidian University | Li S.,Key Laboratory of Intelligent Perception and Image Understanding | Wang S.,Xidian University
Proceedings of the IEEE International Conference on Software Engineering and Service Sciences, ICSESS | Year: 2014

Virtual machine (VM) placement is a key technologyto improve data center efficiency. Most works consider VM placement problem only with respect to physical machine(PM) or network resource optimization. However, efficient VM placement should be implemented by joint optimization of above two aspects. In this paper, a multi-objective VM placement model to minimize the number of active PMs, minimize communication traffic and balance multi-dimensional resource use simultaneously within the data center is proposed. The improved evolutionary multi-objective algorithm: NS-GGA is also designed to tackle this problem, which incorporates the fast non-dominated sorting of NSGA-II into the Grouping Genetic Algorithms. The simulation results show that, in most cases, our model and algorithm gains significantly in all aspects and yields better solutions compared to the existing methods. © 2014 IEEE.


Yang S.,Key Laboratory of Intelligent Perception and Image Understanding | Han Y.,Key Laboratory of Intelligent Perception and Image Understanding | Zhang X.,Key Laboratory of Intelligent Perception and Image Understanding
Proceedings of the International Joint Conference on Neural Networks | Year: 2012

In this paper, we propose a sparse kernel representation classification algorithm (SKRC) for images classification and recognition. The training dictionary is composed by labeled samples directly, and both training dictionary and testing sample are mapped into feature space from original sample space by the sparse kernel which employs the "center" samples matrix constructed by a method similar to k-means clustering. Then in the feature space, the basic sparse representation based classification method is employed. We test our proposed algorithm on some different public database, and the results show that our proposed method can achieve higher classification accuracy without much time consumed. © 2012 IEEE.

Loading Key Laboratory of Intelligent Perception and Image Understanding collaborators
Loading Key Laboratory of Intelligent Perception and Image Understanding collaborators