Advanced Materials Research Vols. 760-762

Abstract: Passive millimeter-wave (PMMW) imaging is an effective technique for concealed objects detection. A PMMW imaging system optimized for corresponding with indoor surroundings is developed and tested. The metal is deposited on the surface of carbon fiber by chemical plating process and the temperature contrast between the human body and concealed objects can be improved by adding the metal deposited carbon fiber to concealed objects. An image enhancement algorithm is proposed on the basis of the combination of wavelet transformation and top-hit transformation, and the experiment results demonstrate that the algorithm proposed is highly efficient for the indoor passive millimeter-wave images. Furthermore, Lucy-Richardson iterative, restricted least square algorithm and user-defined palliative module filtering are combined to process the detecting images, and then the edge detection can be effectively realized using Canny arithmetic operator.

Abstract: In this paper, we present a new approach to improve extracting accuracy of impervious surfaces. One Landsat TM image of Taiyuan city, Shanxi Province of China was used. After doing test work and analyzing using optimum bands analysis, principal component analysis, and normalized difference impervious surface index, we present the method, optimized band combination. Both unsupervised and supervised classification methods were used to classify the original image, principal component analysis image, normalized difference impervious surface index image, and optimized band combination images we present. The accuracies result of these classifications were assessed by using 256 randomly selected sampling points, and it was found that the overall accuracy the accuracy of optimized band combination method can be reach 87.72%, with the Kappa statistic of 0.85 in impervious surface extraction, it was better than other three methods can get.

Abstract: Medical image plays an important role in the assist doctors in the diagnosis and treatment of diseases. For the medical image, the further analysis and diagnosis of the target area is based on image segmentation. There are many different kinds of image segmentation algorithms. In this paper, image segmentation algorithms are divided into classical image segmentation algorithms and segmentation methods combined with certain mathematical tools, including threshold segmentation methods, image segmentation algorithms based on the edge, image segmentation algorithms based on the region, image segmentation algorithms based on artificial neural network technology, image segmentation algorithms based on contour model and image segmentation algorithm based on statistical major segmentation algorithm and so on. Finally, the development trend of medical image segmentation algorithms is discussed.

Abstract: Feature matching used in SIFT methods to extracting feature points, is the key of image mosaic. The advantages of SIFT is one of the most robust and the widely used image matching algorithms based on local features, which ensure the good mosaic image and reliable result. In this paper, we proposed a new image mosaic technology based on simple, which can solve massive calculation and long processing time produced by SIFT. The first step from this method is that 32 vectors, from the feature points, reduce to 16 vectors, and the feature points with 16 vectors construction a pot set, which is the space transformation matrix. And space transformation, the transformation is the mapping of the two images. The processed image is edge processed to eliminate image edge ambiguity and the right results is also rests on cluster .The stitched image is achieved thorough it. Experimental results show that the proposed algorithm is effective and comparing favorable with existing techniques, Practical application shows that this method also proposes a reliable parameter estimation method, and the result is reliable to stitching a large image. At the same time, the processing time is faster than SIFT.

Abstract: According to the characteristic of SAR image containing the multiplicative speckle, a kind of neural network image compression algorithm in wavelet domain was proposed. Wavelet transform can well reflect the characteristics of human vision, but the neural network has self-learning, adaptive, robust, highly parallel processing ability and generalization ability. The wavelet and neural network together in SAR image compression compared with other encoding methods, has obvious advantage. Compared with block-DCT algorithm and sub-band DWT method, this algorithm preserves more advantage in speckle reduction and image details keeping, the compressed image with visual features the best.

Abstract: Content based image retrieval (CBIR) is an essential task in many applications. Color based methods have received much attention in past years, since color could serve efficiently for image retrieval, especially in the case of large database. However, there are two main drawbacks for color based image retrieval methods. Firstly, color based methods are not suitable for similar scenes under different illumination conditions, because color is sensitive to illumination. Secondly, existing approaches usually employ image descriptors with large size, which makes the approach unsuitable for real-time application. To overcome drawbacks mentioned above, an adaptive image retrieval method has been proposed, which integrates the color invariant with the spatial information about images. Different from previous methods, the quantization of the color space has not been manually determined. Instead, it has been decided according to the content of image, using an adaptive clustering technique. Therefore, the size of image descriptor is very small. In the proposed method, feature maps for images have been firstly established, which consist of color invariants. And then the Markov chain model has been employed to capture color information and spatial features. Finally, similar images are retrieved based on two-stage weighted distance. Experimental results show that the proposed method has improved simplicity and compactness of color based image retrieval methods, without the loss of efficiency and robustness.

Abstract: When the number of labeled training samples is very small, the sample information we can use would be very little. Because of this, the recognition rates of some traditional image recognition methods are not satisfactory. In order to use some related information that always exist in other databases, which is helpful to feature extraction and can improve the recognition rates, we apply multi-task learning to feature extraction of images. Our researches are based on transferring the projection transformation. Our experiments results on the public AR, FERET and CAS-PEAL databases demonstrate that the proposed approaches are more effective than the general related feature extraction methods in classification performance.

Abstract: As one of the most popular research topics, sparse representation (SR) technique has been successfully employed to solve face recognition task. Though current SR based methods prove to achieve high classification accuracy, they implicitly assume that the losses of all misclassifications are the same. However, in many real-world face recognition applications, this assumption may not hold as different misclassifications could lead to different losses. Driven by this concern, in this paper, we propose a cost-sensitive sparsity preserving projections (CSSPP) for face recognition. CSSPP considers the cost information of sparse representation while calculating the sparse structure of the training set. Then, CSSPP employs the sparsity preserving projection method to achieve the projection transform and keeps the sparse structure in the low-dimensional space. Experimental results on the public AR and FRGC face databases are presented to demonstrate that both of the proposed approaches can achieve high recognition rate and low misclassification loss, which validate the efficacy of the proposed approach.

Abstract: Color Image Recognition is one of the most important fields in Pattern Recognition. Both Multi-set canonical correlation analysis and Kernel method are important techniques in the field of color image recognition. In this paper, we combine the two methods and propose one novel color image recognition approach: color image kernel canonical correlation analysis (CIKCCA). Color image kernel canonical correlation analysis is based on the theory of multi-set canonical correlation analysis and extracts canonical correlation features among the color image components. Then fuse the features of the color image components in the feature level, which are used for classification and recognition. Experimental results on the FRGC-v2 public color image databases demonstrate that the proposed approach acquire better recognition performance than other color recognition methods.

Abstract: In the field of face recognition, how to extract effective nonlinear discriminative features is an important research topic. In this paper, we propose a new kernel orthogonal projection analysis approach. We obtain the optimal nonlinear projective vector which can differentiate one class and its adjacent classes, by using the Fisher criterion and constructing the specific between-class and within-class scatter matrices in kernel space. In addition, to eliminate the redundancy among projective vectors, our approach makes every projective vector satisfy locally orthogonal constraints by using the corresponding class and part of its most adjacent classes. Experimental results on the public AR and CAS-PEAL face databases demonstrate that the proposed approach outperforms several representative nonlinear projection analysis methods.