Showing papers on "Contourlet published in 2012"

NSCT-based multimodal medical image fusion using pulse-coupled neural network and modified spatial frequency

Abstract: In this article, a novel multimodal medical image fusion (MIF) method based on non-subsampled contourlet transform (NSCT) and pulse-coupled neural network (PCNN) is presented. The proposed MIF scheme exploits the advantages of both the NSCT and the PCNN to obtain better fusion results. The source medical images are first decomposed by NSCT. The low-frequency subbands (LFSs) are fused using the ‘max selection’ rule. For fusing the high-frequency subbands (HFSs), a PCNN model is utilized. Modified spatial frequency in NSCT domain is input to motivate the PCNN, and coefficients in NSCT domain with large firing times are selected as coefficients of the fused image. Finally, inverse NSCT (INSCT) is applied to get the fused image. Subjective as well as objective analysis of the results and comparisons with state-of-the-art MIF techniques show the effectiveness of the proposed scheme in fusing multimodal medical images.

Recognition of driving postures by contourlet transform and random forests

Abstract: An efficient feature extraction approach for driving postures from a video camera, which consists of Homomorphic filtering, skin-like regions segmentation and contourlet transform (CT), was proposed. With features extracted from a driving posture dataset created at Southeast University (SEU), holdout and cross-validation experiments on driving posture classification were then conducted using random forests (RF) classifier. Compared with a number of commonly used classification methods including linear perceptron classifier, k -nearest-neighbour classifier and multilayer perceptron (MLP) classifier, the experiments results showed that the RF classifier offers the best classification performance among the four classifiers. Among the four predefined classes, that is, grasping the steering wheel, operating the shift gear, eating and talking on a cellular phone, the class of eating is the most difficult to classify. With RF classifier, the classification accuracies of eating are over 88% in holdout and cross-validation experiments, thus demonstrating the effectiveness of the proposed feature extraction method and the importance of RF classifier in automatically understanding and characterising driver%s behaviours towards human-centric driver assistance systems.

Image fusion by pulse couple neural network with shearlet

Abstract: The shearlet representation forms a tight frame which decom- poses a function into scales and directions, and is optimally sparse in representing images with edges. An image fusion method is proposed based on the shearlet transform. Firstly, transform the image A and image B by the shearlets. Secondly, a pulse couple neural network (PCNN) is used for the frequency subbands, which uses the number of output pulses from the PCNN's neurons to select fusion coefficients. Finally, an inverse shearlet transform is applied on the new fused coeffi- cients to reconstruct the fused image. Some experiments are performed in images such as multi-focus images, multi-sensor images, medical images and multispectral images comparing the proposed algorithm with the wavelet, contourlet and nonsubsampled contourlet method based on the PCNN. The experimental results show that the proposed algorithm can not only extract more important visual information from source images, but also effectively avoid the introduction of artificial information. It signifi- cantly outperforms the traditional multiscale transform image fusion meth- ods in terms of both visual quality and objective evaluation criteria such as MI and Q AB∕F. © 2012 Society of Photo-Optical Instrumentation Engineers (SPIE). (DOI: 10

Contourlet hidden Markov Tree and clarity-saliency driven PCNN based remote sensing images fusion

Abstract: In this paper a new Contourlet hidden Markov Tree (CHMT) and clarity-saliency driven Pulse Coupled Neural Network (PCNN) based fusion approach is proposed for remote sensing images fusion. Considering the failure of wavelet in representing the geometry of image edges and contours, we firstly use Contourlet to provide an efficient and flexible multiscale, multidirectional and anisotropy representation of remote sensing images, and then use CHMT model to describe the statistics of Contourlet coefficients, by taking the dependencies across scales, locations and directions into account. Because CHMT can span several adjacent directional subbands in the finer scales, which has similar statistical characteristic to scale, CHMT model can give more accurate description of images and has lower computational complexity than wavelet hidden Markov Tree (WHMT). Training the Contourlet coefficients of registered multisource images using Expectation Maximum (EM) algorithm, one can obtain the model parameters which is used to update the Contourlet coefficients. Low-frequency subbands are fused by the magnitude maximum rule. For the fusion of high-frequency directional subband, a PCNN is constructed based on the phenomena of synchronous pulse bursts in the animal visual cortexes, and the linking strength of each neuron in PCNN is determined by a new clarity-saliency measure of the subband images. New fire mapping images are obtained for each high-frequency subband taking part in the fusion. Some experiments are taken on some remote sensing images came from the USA Airborne Multisensor Pod System (AMPS) program, and the results show the superiorities of our proposed method to WHMT and Contourlets, both in subjective evaluation, implementation speed and some numerical guidelines.

On the choice of Compressed Sensing priors and sparsifying transforms for MR image reconstruction: An experimental study

Abstract: Recently Compressed Sensing (CS) based techniques are being used for reconstructing magnetic resonance (MR) images from partially sampled k-space data. CS based reconstruction techniques can be categorized into three categories based on the objective function: (i) synthesis prior, (ii) analysis prior and (iii) mixed (analysis+synthesis) prior. Each of these can be further subdivided into convex and non-convex forms. There is also a wide choice available for the sparsifying transforms, viz. Daubechies wavelets (orthogonal and redundant), fractional spline wavelet (orthogonal), complex dualtree wavelet (redundant), contourlet (redundant) and finite difference (redundant). Previous studies in MR image reconstruction have used a various combinations of objective functions (priors) and sparsifying transforms; and each of these studies claimed the superiority of their method over others. In this work, we will review and evaluate the popular MR image reconstruction techniques and show that analysis prior with complex dualtree wavelets yields the best reconstruction results. We have evaluated our experimental results on real data. The metric for quantitative evaluation is the Normalized Mean Squared Error. Our qualitative evaluation is based both on the reconstructed and the difference images. The other significant contribution of this paper is the development of convex and non-convex versions of synthesis, analysis and mixed prior algorithms from a uniform majorization-minimization framework. The algorithms are compared with a state-of-the-art CS based techniques; the proposed ones have better reconstruction accuracy and are only fractionally slow. The algorithms that are derived in this paper are all efficient first order algorithms that are easy to implement.

Effective Management of Medical Information Through A Novel Blind Watermarking Technique

Abstract: Medical Data Management (MDM) domain consists of various issues of medical information like authentication, security, privacy, retrieval and storage etc. Medical Image Watermarking (MIW) techniques have recently emerged as a leading technology to solve the problems associated with MDM. This paper proposes a blind, Contourlet Transform (CNT) based MIW scheme, robust to high JPEG and JPEG2000 compression and simultaneously capable of addressing a range of MDM issues like medical information security, content authentication, safe archiving and controlled access retrieval etc. It also provides a way for effective data communication along with automated medical personnel teaching. The original medical image is first decomposed by CNT. The Low pass subband is used to embed the watermark in such a way that enables the proposed method to extract the embedded watermark in a blind manner. Inverse CNT is then applied to get the watermarked image. Extensive experiments were carried out and the performance of the proposed scheme is evaluated through both subjective and quantitative measures. The experimental results and comparisons, confirm the effectiveness and efficiency of the proposed technique in the MDM paradigm.

Multitemporal Image Change Detection Using a Detail-Enhancing Approach With Nonsubsampled Contourlet Transform

Abstract: In this letter, we propose an unsupervised approach for change detection in multitemporal satellite images based on a novel detail-enhancing algorithm. The multitemporal source images are first used to generate the difference image, which is decomposed into low-pass approximation and high-pass directional subbands by the nonsubsampled contourlet transform. The coefficients from the directional subbands are fused at intrascale and interscale to extract the meaningful details of the difference image. After that, the extracted details are injected into one base image selected from the approximation subbands, which results in a detail-enhanced difference image. For each pixel in the enhanced difference image, a dimension-reduced feature vector is created using the principal component analysis (PCA). The final change detection map is achieved by clustering the feature vectors using a PCA-guided k-means algorithm into “changed” and “unchanged” classes. Experimental results demonstrate the superior performance of the proposed approach compared with several well-known change detection techniques.

A new SVM-based image watermarking using Gaussian-Hermite moments

Abstract: Geometric attack is known as one of the most difficult attacks to resist, for it can desynchronize the location of the watermark and hence causes incorrect watermark detection. It is a challenging work to design a robust image watermarking scheme against geometric attacks. Based on the support vector machine (SVM) and Gaussian-Hermite moments (GHMs), we propose a robust image watermarking algorithm in nonsubsampled contourlet transform (NSCT) domain with good visual quality and reasonable resistance toward geometric attacks in this paper. Firstly, the NSCT is performed on original host image, and corresponding low-pass subband is selected for embedding watermark. Then, the selected low-pass subband is divided into small blocks. Finally, the digital watermark is embedded into host image by modulating adaptively the NSCT coefficients in small block. The main steps of digital watermark detecting procedure include: (1) some low-order Gaussian-Hermite moments of training image are computed, which are regarded as the effective feature vectors; (2) the appropriate kernel function is selected for training, and a SVM training model can be obtained; (3) the watermarked image is corrected with the well trained SVM model; (4) the digital watermark is extracted from the corrected watermarked image. Experimental results show that the proposed image watermarking is not only invisible and robust against common image processing operations such as filtering, noise adding, JPEG compression, etc., but also robust against the geometric attacks.

Improved contourlet-based steganalysis using binary particle swarm optimization and radial basis neural networks

Abstract: Steganography is the science of hiding information in a media such as video, image or audio files. On the other hand, the aim of steganalysis is to detect the presence of embedded data in a given media. In this paper, a steganalysis method is presented for the colored joint photographic experts group images in which the statistical moments of contourlet transform coefficients are used as the features. In this way, binary particle swarm optimization algorithm is also employed as a closed-loop feature selection method to select the efficient features in tandem with improvement of the detection rate. Nonlinear support vector machine and two variants of radial basis neural networks, i.e., radial basis function and probabilistic neural network, are used as the classification tools and their performance is compared in detecting the stego and clean images. Experimental results show that even for low embedding rates, the detection accuracy of the proposed method is more than 80% along with 30% reduction in the size of feature set.

Compressed sensing theory-based reconstruction method of magnetic resonance image

Abstract: The invention provides a compressed sensing theory-based reconstruction method of a magnetic resonance random sampled K space data image. The reconstruction method applies a contourlet conversion and iterative soft thresholding method to realize reconstruction of a magnetic resonance image. The method comprises the following steps: collecting K space data in a magnetic resonance image scanner according to a preset observation matrix phi to generate a measurement value, and keeping y; acquiring y from a coil of the magnetic resonance image scanner, and transmitting y to a computer; and finallyconstructing a same phi, constructing any orthogonal transformation psi, and recovering from y by adopting a compressed sensing theory-based magnetic resonance random sampled K space data image reconstruction method according to reconstruction. According to the method, scanning time is saved, quick imaging is realized, high-quality reliable image information is provided to medical nuclear magnetic resonance imaging detection, and solid theoretical and practical foundation is established for further development and large-scale popularization and application of the medical imaging detection technology.

Liver tumor diagnosis by gray level and contourlet coefficients texture analysis

Abstract: Computed tomography image based Computer Aided Diagnosis (CAD) could be crucially important in supporting liver cancer diagnosis. An effective approach to realize a CAD system for this purpose is described in this work. The CAD system employs automatic tumor segmentation, texture feature extraction and characterization into malignant and benign tumors. A Region of Interest (ROI) cropped from the automatically segmented tumor by confidence connected region growing and alternative fuzzy c means clustering is decomposed using multiresolution and multidirectional contourlet transform to obtain contourlet coefficients. Both first order statistic and second order statistic features are extracted from the gray level and contourlet detail coefficients. The extracted feature sets are classified by a Probabilistic Neural Network (PNN) classifier into benign and malignant. The system is evaluated by using different performance measures and the results indicate that the contourlet coefficient texture is effective for classifying malignant and benign liver tumors from abdominal CT imaging.

Bayesian Texture Classification Based on Contourlet Transform and BYY Harmony Learning of Poisson Mixtures

Abstract: As a newly developed 2-D extension of the wavelet transform using multiscale and directional filter banks, the contourlet transform can effectively capture the intrinsic geometric structures and smooth contours of a texture image that are the dominant features for texture classification. In this paper, we propose a novel Bayesian texture classifier based on the adaptive model-selection learning of Poisson mixtures on the contourlet features of texture images. The adaptive model-selection learning of Poisson mixtures is carried out by the recently established adaptive gradient Bayesian Ying-Yang harmony learning algorithm for Poisson mixtures. It is demonstrated by the experiments that our proposed Bayesian classifier significantly improves the texture classification accuracy in comparison with several current state-of-the-art texture classification approaches.

Improved Image Fusion Method Based on NSCT and Accelerated NMF

Abstract: In order to improve algorithm efficiency and performance, a technique for image fusion based on the Non-subsampled Contourlet Transform (NSCT) domain and an Accelerated Non-negative Matrix Factorization (ANMF)-based algorithm is proposed in this paper. Firstly, the registered source images are decomposed in multi-scale and multi-direction using the NSCT method. Then, the ANMF algorithm is executed on low-frequency sub-images to get the low-pass coefficients. The low frequency fused image can be generated faster in that the update rules for W and H are optimized and less iterations are needed. In addition, the Neighborhood Homogeneous Measurement (NHM) rule is performed on the high-frequency part to achieve the band-pass coefficients. Finally, the ultimate fused image is obtained by integrating all sub-images with the inverse NSCT. The simulated experiments prove that our method indeed promotes performance when compared to PCA, NSCT-based, NMF-based and weighted NMF-based algorithms.

Image denoising based on the symmetric normal inverse Gaussian model and non-subsampled contourlet transform

Abstract: In this study, an adaptive image denoising method is proposed based on the symmetric normal inverse Gaussian (SNIG) model and the non-subsampled contourlet transform (NSCT). In the framework of Bayesian maximum a posteriori estimation, the problem of denoising is reduced to a procedure of thresholding. A novel strategy is then proposed to determine the threshold that is not only adaptive to different directions and scales, but also able to take into considerations the scale-to-scale difference in the contribution of the NSCT coefficients to the noise. The experimental results in different kinds of sample images show that the authors' method can not only result in higher peak-signal-to-noise ratio values, but also have better visual effects in reduced processing artefacts and preserved edges.

Feature extraction based on contourlet transform and its application to surface inspection of metals

Abstract: Surface defects that affect the quality of metals are an important factor. Machine vision systems commonly perform surface inspection, and feature extraction of defects is essential. The rapidity and universality of the algorithm are two crucial issues in actual application. A new method of feature extraction based on contourlet transform and kernel locality preserving projections is proposed to extract sufficient and effective features from metal surface images. Image information at certain direction is important to recognition of defects, and contourlet transform is introduced for its flexible direction setting. Images of metal surfaces are decomposed into multiple directional subbands with contourlet transform. Then features of all subbands are extracted and combined into a high-dimensional feature vector, which is reduced to a low-dimensional feature vector by kernel locality preserving projections. The method is tested with a Brodatz database and two surface defect databases from industrial surface-inspection systems of continuous casting slabs and aluminum strips. Experimental results show that the proposed method performs better than the other three methods in accuracy and efficiency. The total classification rates of surface defects of continuous casting slabs and aluminum strips are up to 93.55% and 92.5%, respectively.

An effective similarity measure via genetic algorithm for Content-Based Image Retrieval with extensive features

Abstract: With the aid of image content, the relevant images can be extracted from the image in the Content Based Image Retrieval (CBIR) system. Concise feature sets limit the retrieval efficiency, to eliminate this shape, colour, texture and contourlet features are extracted. For retrieving relevant images, the optimisation technique Genetic Algorithm (GA) is utilised and for similarity measure Squared Euclidean Distance (SED) is utilised for comparing query image featureset and database image featureset. Hence, from GA based similarity measure, relevant images are retrieved and evaluated by querying different images.

Multi-focus image fusing method based on dual-channel PCNN (Pulse Coupled Neural Network)

Abstract: The invention discloses a multi-focus image fusing method based on a dual-channel PCNN (Pulse Coupled Neural Network), which belongs to the technical field of image processing. The method comprises the following steps of: performing NSCT (Non-Subsampled Contourlet Transform) on two images respectively to obtain a plurality of sub-images of different frequencies; fusing by correspondingly adopting the dual-channel PCNN, and determining each band pass sub-band coefficient of a fused image; and performing reverse NSCT to obtain the fused image. Due to the adoption of the multi-focus image fusing method, the defects of the conventional multi-focus image fusing method are overcome, and the fusing effect is improved.

A Comparative Study in Wavelets, Curvelets and Contourlets as Denoising Biomedical Images

Abstract: A special member of the emerging family of multi scale geometric transforms is the contourlet transform which was developed in the last few years in an attempt to overcome inherent limitations of traditional multistage representations such as curvelets and wavelets. The biomedical images were denoised using firstly wavelet than curvelets and finally contourlets transform and results are presented in this paper. It has been found that contourlets transform outperforms the curvelets and wavelet transform in terms of signal noise ratio

An improved boosting algorithm and its application to facial emotion recognition

Abstract: This paper develops a regularized discriminant analysis (RDA)-based boosting algorithm, and its application of the facial emotion recognition. The RDA-based boosting algorithm uses RDA as a learning rule in the boosting algorithm. The RDA combines strengths of linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA). It solves small sample size and ill-posed problems suffered from QDA and LDA through a regularization technique. Additionally, this study uses a particle swarm optimization algorithm to estimate optimal parameters in RDA. In this work, the proposed RDA-based boosting is used in the facial emotion recognition, and achieves a good performance. In the facial emotion recognition, contourlet features are extracted and followed by an entropy criterion to select the informative contourlet features which is a subset of informative and non-redundant contourlet features. Experiment results demonstrate that the proposed RDA-based boosting can accurately and robustly recognize facial emotions.

The Translation Invariant Wavelet-based Contourlet Transform for Image Denoising

Abstract: A new method of image denoising using wavelet-based contourlet transform (WBCT) is proposed. Due to the lack of translation invariance of WBCT, image denoising by means of WBCT would lead to Gibbs-like phenomena. In the paper,cycle spinning-based technique is applied to develop translation invariant WBCT denoising scheme. Many simulation experiments with images contaminated by additive white Gaussian noise demonstrate that the performance of the proposed approach substantially surpasses that of previously wavelets methods using the cycle spinning both visually and in terms of the PSNR values, especially for the images that include mostly fine textures and contours.

A multi-resolution watermarking based on contourlet transform using SVD and QR decomposition

Abstract: A novel image watermarking algorithm is developed which is based on Contourlet Transform (CT) and Matrix Factorization methods such as Singular Value Decomposition (SVD) technique and QR factorization method. In the proposed scheme, the original image and the watermark image are transformed into subbands by Contourlet Transform. The lowest frequency coefficients of the CT decomposed original image are further factorized using SVD. The image coefficients of the lowest frequency of watermark after CT are decomposed by QR factorization. The QR decomposed coefficients are further factorized by SVD technique. The coefficients of watermark image thus obtained are embedded into the SVD factorized original image values. Inverse SVD and inverse CT are then applied on the modified coefficients of the original image to get the watermarked image. The performance of watermarking algorithm is improved by the combination of CT, SVD and QR factorization. Experimental results show that the proposed algorithm is better than the other contourlet based watermarking technique. It is robust and imperceptible against common signal processing attacks such as scaling, compression and filtering.

Face recognition based on nonsubsampled contourlet transform and block-based kernel Fisher linear discriminant

Abstract: Face representation, including both feature extraction and feature selection, is the key issue for a successful face recognition system. In this paper, we propose a novel face representation scheme based on nonsubsampled contourlet transform (NSCT) and block-based kernel Fisher linear discriminant (BKFLD). NSCT is a newly developed multiresolution analysis tool and has the ability to extract both intrinsic geometrical structure and directional information in images, which implies its discriminative potential for effective feature extraction of face images. By encoding the the NSCT coefficient images with the local binary pattern (LBP) operator, we could obtain a robust feature set. Furthermore, kernel Fisher linear discriminant is introduced to select the most discriminative feature sets, and the block-based scheme is incorporated to address the small sample size problem. Face recognition experiments on FERET database demonstrate the effectiveness of our proposed approach.

Fusion method of SAR (Synthetic Aperture Radar) images and visible light images on the basis of NSCT (Non Subsampled Contourlet Transform)

Abstract: The invention relates to a fusion method of SAR (Synthetic Aperture Radar) images and visible light images on the basis of NSCT (Non Subsampled Contourlet Transform). The fusion method is characterized by comprising the following steps of: firstly, carrying out NSCT decomposition on the SAR images and the visible light images respectively; then adopting different fusion rules to carry out fusion treatment on NSCT low-frequency and high-frequency subband coefficients, wherein according to the decomposition coefficient characteristics of noise and signals in an NSCT domain, carrying out hard-threshold denoising on the NSCT high-frequency subband coefficient of the SAR images under the maximum decomposition scale, then respectively adopting different fusion rules to carry out fusion processing on the NSCT high-frequency subband coefficient under the maximum decomposition scale and other decomposition scales by adopting the coefficients with threshold processing as the basis; and finally,carrying out NSCT reverse transformation on the fused NSCT coefficients and obtaining fused images. The fusion method takes denoising as the basis of the fusion rule design, considers noise suppression while fusion treatment is carried out, is simple and easy to operate, can be used for obtaining a good fusion effect and is especially more applicable to the SAR images and the visible light imageswith serious spot and noise pollution.

Edge preserving image fusion based on contourlet transform

Abstract: Image fusion is an emerging area of research having a number of applications in medical imaging, remote sensing, satellite imaging, target tracking, concealed weapon detection and biometrics. In the present work, we have proposed a new edge preserving image fusion method based on contourlet transform. As contourlet transform has high directionality and anisotropy, it gives better image representation than wavelet transforms. Also contourlet transform represents salient features of images such as edges, curves and contours in better way. So it is well suited for image fusion. We have performed experiments on several image data sets and results are shown for two datasets of multifocus images and one dataset of medical images. On the basis of experimental results, it was found that performance of proposed fusion method is better than wavelet transform (Discrete wavelet transform and Stationary wavelet transform) based image fusion methods. We have verified the goodness of the proposed fusion algorithm by well known image fusion measures (entropy, standard deviation, mutual information (MI) and $Q_{AB}^{F}$). The fusion evaluation parameters also imply that the proposed edge preserving image fusion method is better than wavelet transform (Discrete wavelet transform and Stationary wavelet transform) based image fusion methods.