Detail Subbands Research Articles

Image denoising using SVM classification in nonsubsampled contourlet transform domain

For image denoising, the main challenge is how to preserve the information-bearing structures such as edges and textures to get satisfactory visual quality when improving the signal-to-noise-ratio (SNR). Edge-preserving image denoising has become a very intensive research topic. In this paper, we propose an image denoising using support vector machine (SVM) classification in nonsubsampled contourlet transform (NSCT) domain. Firstly, the noisy image is decomposed into different subbands of frequency and orientation responses using the NSCT. Secondly, the feature vector for a pixel in a noisy image is formed by the spatial regularity in NSCT domain, and the least squares support vector machine (LS-M) model is obtained by training. Then the NSCT detail coefficients are divided into two classes (edge-related coefficients and noise-related ones) by LS-SVM training model. Finally, the detail subbands of NSCT coefficients are denoised by using shrink method, in which the adaptive Bayesian threshold is utilized. Extensive experimental results demonstrate that our method can obtain better performances in terms of both subjective and objective evaluations than those state-of-the-art denoising techniques. Especially, the proposed method can preserve edges very well while removing noise.

Image resampling detection based on texture classification

This study presents a method for resampling detection. By combining texture analysis with resampling detection, the task of resampling detection is considered as a texture classification problem. In other words, the influence of resampling operations on a raw single-sampled image is viewed as an alteration of the image texture in a fine scale. First, local linear transform is used to obtain textural detail sub-bands. A 36-D feature vector is then extracted from the normalized characteristic function moments of textural detail sub-bands to train a support vector machine classifier. Finally, experimental results are reported on three databases, with each having almost 10,000 images. Comparison with the previous study reveals that the proposed method is effective for resampling detection. In addition, extensive experiments on cover and stego bitmap images illustrate that the proposed method is essential for constructing accurate targeted and blind steganalysis methods for heterogeneous images, raw single-sampled images, and images resampled at different scales.

Wavelet-based Despeckling of Medical Ultrasound Images

This paper presents a new wavelet-domain technique for despeckling of medical ultrasound (US) images for improved clinical diagnosis. The method uses the generalized Gaussian distribution and generalized gamma distribution to model the image and the speckle, respectively, in the detailed sub-bands of wavelet decomposition of the log-transformed US image. Combining these, a priori distributions with the Bayesian maximum a posteriori criterion, shrinkage estimators are derived for processing the wavelet coefficients of the detail sub-bands. The visual comparison of despeckled US images and the higher values of quality metrics indicate that the new method suppresses the speckle noise well while preserving the texture and organ surfaces.

Mathematical Methods in Biomedical Imaging

Mathematical Methods in Biomedical Imaging

A Fusion Algorithm for GFP Image and Phase Contrast Image of Arabidopsis Cell Based on SFL-Contourlet Transform

A hybrid multiscale and multilevel image fusion algorithm for green fluorescent protein (GFP) image and phase contrast image of Arabidopsis cell is proposed in this paper. Combining intensity-hue-saturation (IHS) transform and sharp frequency localization Contourlet transform (SFL-CT), this algorithm uses different fusion strategies for different detailed subbands, which include neighborhood consistency measurement (NCM) that can adaptively find balance between color background and gray structure. Also two kinds of neighborhood classes based on empirical model are taken into consideration. Visual information fidelity (VIF) as an objective criterion is introduced to evaluate the fusion image. The experimental results of 117 groups of Arabidopsis cell image from John Innes Center show that the new algorithm cannot only make the details of original images well preserved but also improve the visibility of the fusion image, which shows the superiority of the novel method to traditional ones.

Wavelet-Based Dynamic Texture Classification Using Gumbel Distribution

Dynamic texture classification has attracted growing attention. Characterization of a dynamic texture is vital to address the classification problem. This paper proposes a dynamic texture descriptor based on the dual-tree complex wavelet transform and the Gumbel distribution. The method takes out the median values of coefficient magnitudes in each nonoverlapping block of a detail subband and models them with the Gumbel distribution. The classification is realized by comparing the similarity between the estimated distributions of all detail subbands. The experimental results on the benchmark dynamic texture database demonstrate better histogram fitting and promising classification performance of the dynamic texture descriptor compared with the current existing methods.

Analysis and Denoising of Hyperspectral Remote Sensing Image in the Curvelet Domain

A new denoising algorithm is proposed according to the characteristics of hyperspectral remote sensing image (HRSI) in the curvelet domain. Firstly, each band of HRSI is transformed into the curvelet domain, and the sets of subband images are obtained from different wavelength of HRSI. And then the detail subband images in the same scale and same direction from different wavelengths of HRSI are stacked to obtain new 3-D datacubes of the curvelet domain. Again, the characteristics analysis of these 3-D datacubes is performed. The analysis result shows that each new 3-D datacube has the strong spectral correlation. At last, due to the strong spectral correlation of new 3-D datacubes, the multiple linear regression is introduced to deal with these new 3-D datacubes in the curvelet domain. The simulated and the real data experiments are performed. The simulated data experimental results show that the proposed algorithm is superior to the compared algorithms in the references in terms of SNR. Furthermore, MSE and MSSIM in each band are utilized to show that the proposed algorithm is superior. The real data experimental results show that the proposed algorithm effectively removes the common spotty noise and the strip noise and simultaneously maintains more fine features during the denoising process.

Hierarchical wavelet packet fuzzy inference system for pattern classification and system identification

This study presents a hierarchical Takagi–Sugeno–Kang type fuzzy system called hierarchical wavelet packet fuzzy inference system. In the proposed method, wavelet packet transform is applied on the input data to produce approximation and detail sub-bands of the input data and the output is used as the input vector of the proposed network. This network uses a hierarchical structure same as wavelet packet decomposition tree, in which adaptive network-based fuzzy inference system is used as sub-model. Also, gradient descent algorithm is chosen for training the parameters of antecedent and conclusion parts of the sub-models. In order to evaluate the capability of the proposed method, its applications in pattern classification, system identification and time-series prediction have been studied. The results show that the proposed method performs better than the other conventional models.

Image Watermarking Algorithm Based on Multiobjective Ant Colony Optimization and Singular Value Decomposition in Wavelet Domain

We present a new optimal watermarking scheme based on discrete wavelet transform (DWT) and singular value decomposition (SVD) using multiobjective ant colony optimization (MOACO). A binary watermark is decomposed using a singular value decomposition. Then, the singular values are embedded in a detailed subband of host image. The trade-off between watermark transparency and robustness is controlled by multiple scaling factors (MSFs) instead of a single scaling factor (SSF). Determining the optimal values of the multiple scaling factors (MSFs) is a difficult problem. However, a multiobjective ant colony optimization is used to determine these values. Experimental results show much improved performances of the proposed scheme in terms of transparency and robustness compared to other watermarking schemes. Furthermore, it does not suffer from the problem of high probability of false positive detection of the watermarks.

Multivariate Generalized Gamma Distribution for Content Based Image Retrieval

This paper deals with the joint modeling of color textures in the context of Content Based Image Retrieval (CBIR). We propose a generic multivariate model based on the Generalized Gamma distribution to describe the marginal behavior of texture wavelet detail subbands. Then the information of dependence across color components is incorporated in the modeling process using the Gaussian copula. The multivariate Generalized Gamma distribution ( D MG ) is advantageous in term of flexibility when compared with other joint models. As similarity measure, we propose a closed-form of the Kullback-leibler (KL) divergence between two Ds MG . Performances of the CBIR system show the superiority of the proposed model over a variety of multivariate models.

Image textural features for steganalysis of spatial domain steganography

From the texture analysis of image content, we propose a steganalytic method to detect spatial domain steganography in grayscale images. First of all, based on the local linear vectors, which are selected carefully and sensitive to image texture, images are decomposed into several textural detail subbands by the local linear transform (LLT). Then the statistical distribution of the LLT coefficient is modeled by using the generalized Gaussian distribution. Finally, novel textural features of the LLT coefficient histogram and cooccurrence matrix are extracted for steganalyzers implemented by the support vector machine. Extensive experiments are performed on four diverse uncompressed image databases and seven typical spatial domain steganographic algorithms, such as the highly undetectable stego. The results reveal that the proposed scheme is universal for detecting spatial domain steganography. By comparison with other well-known feature sets, our presented feature set offers the best performance under most circumstances.

Score-Level Fusion of Local Spatial, Scale and Directional Features Based Face Recognition Approach for Single Sample Problem

The local appearance based methods have been successfully applied to face recognition and achieved state-of-the-art performance. In this paper we propose a face representation approach which explores the information in the spatial domain along with score-level fusion of directional features in different scales. The local spatial texture features are extracted using the local MinMax binary pattern (LMinMaxBP) approach as well as with the help of the selective local texture feature approach. The dual tree complex wavelet transform (DT-CWT) provides a local multiscale description of images with good directional selectivity, effective edge representation and invariance to shifts and in-plane rotations. It is insensitive to illumination variations and facial expression changes. 2-D dual tree complex wavelet transform is less redundant and computationally efficient. The fusion of local DT-CWT coefficients of detail subbands are used to extract the multi scale facial features with good directional selectivity which improves the face recognition with small sample size in less computation. This score-level fusion approach combines information from different domains to give a good face representation for recognition. Extensive experimental results on FERET, ORL, Yale and Indian Face databases show the significant advantages of the proposed method over the existing ones.

Passive copy move image forgery detection using undecimated dyadic wavelet transform

In this paper, a blind copy move image forgery detection method using undecimated dyadic wavelet transform (DyWT) is proposed. DyWT is shift invariant and therefore more suitable than discrete wavelet transform (DWT) for data analysis. First, the input image is decomposed into approximation (LL1) and detail (HH1) subbands. Then the LL1 and HH1 subbands are divided into overlapping blocks and the similarity between blocks is calculated. The key idea is that the similarity between the copied and moved blocks from the LL1 subband should be high, while that from the HH1 subband should be low due to noise inconsistency in the moved block. Therefore, pairs of blocks are sorted based on high similarity using the LL1 subband and high dissimilarity using the HH1 subband. Using thresholding, matched pairs are obtained from the sorted list as copied and moved blocks. Experimental results show the effectiveness of the proposed method over competitive methods using DWT and the LL1 or HH1 subbands only.

Image Denoising based on Spatial/Wavelet Filter using Hybrid Thresholding Function

n this paper a hybrid denoising algorithm which combines spatial domain bilateral filter and hybrid thresholding function in the wavelet domain is proposed. The wavelet transform is used to decompose the noisy image into its different subbands namely LL, LH, HL, and HH. A two stage spatial bilateral filter is applied. The first stage is applied on the noisy image before wavelet decomposition. This stage will be called a pre- processing stage. The second stage spatial bilateral filtering is applied on the low frequency subband of the decomposed noisy image namely subband LL. This stage will tend to cancel or at least attenuate any residual low frequency noise components. The intermediate stage deal with high frequency noise components by thresholding detail subbands LH, HL, and HH using hybrid thresholding function. The experimental results show that the performance of the proposed denoising algorithm is superior to that of the conventional denoising approach.

Gray Scale Image Compression Based on Wavelet Transform and Linear Prediction

Every year, several terabytes of image data- both medical and non medical- are engendered so that the requisition for image compression is substantiated. In this paper, the correlation properties of wavelets are harnessed in linear predictive coding to compress images. The image is decomposed using a one dimensional wavelet transform. The highest level wavelet transform coefficients and few detail coefficients in every level are retained. Using linear prediction on these coefficients the image is reconstructed. The prediction is done in both the dimensions, so the numbers of coefficients retained in detail subbands are less. With less predictors and samples from the original wavelet coefficients compression can be achieved. The results are appraised in objective and subjective manner with real world and medical images. The results are also verified on ModelFest database.

DWT and DCT based Robust Iris Feature Extraction and Recognition Algorithm for Biometric Personal Identification

iris is one of the most reliable biometric because of its uniqueness, stability and noninvasive nature. Thus it has attracted the attention of biometrics based identification and verification research and development community. In this paper, a new approach of iris image feature extraction technique based on the statistical properties of Discrete Cosine Transform (DCT) domain is proposed. A Canny Edge Detection followed by Hough Transform is used to detect the iris boundaries in the eye's digital image. The two level Discrete Wavelet Transformation (DWT) is applied on the segmented and normalized iris region. Both second level horizontal and vertical detail sub-bands are used for encoding unique iris feature. Each of those frequency sub-bands is divided into 8x8 non-overlapping blocks and DCT is applied to each block. Unique iris features are obtained by comparing the energies containing in corresponding DCT blocks of both the sub-bands. These features extracted are used to generate unique encoded binary image and corresponding unique binary bit stream/code is constructed. In order to reduce the size of the database, this binary bit stream instead of binary image is stored in database for matching purpose. Further to increase the security of the system, the bit stream obtained is first encrypted using the user key obtained from user password and then the encrypted bit pattern template is stored. Experimental results on Iris Database reveal that the proposed iris matching scheme provides results comparable to those of recent methods and is also computationally effective.

Nonlinear filtering based on 3D wavelet transform for MRI denoising

Magnetic resonance (MR) images are normally corrupted by random noise which makes the automatic feature extraction and analysis of clinical data complicated. Therefore, denoising methods have traditionally been applied to improve MR image quality. In this study, we proposed a 3D extension of the wavelet transform (WT)-based bilateral filtering for Rician noise removal. Due to delineating capability of wavelet, 3D WT was employed to provide effective representation of the noisy coefficients. Bilateral filtering of the approximation coefficients in a modified neighborhood improved the denoising efficiency and effectively preserved the relevant edge features. Meanwhile, the detailed subbands were processed with an enhanced NeighShrink thresholding algorithm. Validation was performed on both simulated and real clinical data. Using the peak signal-to-noise ratio (PSNR) to quantify the amount of noise of the MR images, we have achieved an average PSNR enhancement of 1.32 times with simulated data. The quantitative and the qualitative measures used as the quality metrics demonstrated the ability of the proposed method for noise cancellation.

Wavelet-based oblivious image watermarking scheme using genetic algorithm

In this work, a robust and oblivious image watermarking scheme based on discrete wavelet transform (DWT) for copyright protection is presented. The original unmarked image is not required for watermark extraction. Third-level DWT is applied to the original cover image. Third- and second-level horizontal detail sub-band (LH2 and LH3) coefficients are grouped into different blocks. Grouping of the coefficients is done in such a way that each block should contain one coefficient from LH3 sub-band and four coefficients from LH2 sub-band. In each block, the first minimum and the second minimum are identified and modified according to the watermark bit. After watermark insertion, inverse DWT is applied to the sub-bands with modified coefficients to obtain the watermarked image. For watermark extraction, a threshold-based decoder is designed. Embedding and extraction process are characterised with parameters and genetic algorithm is used for parameter optimisation. Optimisation is to maximise the values of peak signal-to-noise ratio of the watermarked image and normalised cross correlation of the extracted watermark. The performance of the proposed scheme is compared with the existing schemes and significant improvement is observed. Experimental results show that, this algorithm is highly robust for many image attacks on the watermarked image.

Joint trellis-coded quantization watermarking for JPEG2000 images

Watermarking in the Joint Photographic Experts Group (JPEG)2000 coding pipeline is investigated in this paper. A joint quantization and watermarking method based on trellis-coded quantization (TCQ) is proposed to reliably embed data during the quantization stage of the JPEG2000 part 2 codec. The central contribution of this work is the use of a single quantization module to jointly perform quantization and watermark embedding at the same time. The TCQ-based watermarking technique allows embedding the watermark in the detail sub-bands of one or more resolution levels except the first one. Watermark recovery is performed after image decompression. The performance of this joint scheme in terms of image quality and robustness against common image attacks was estimated on real images.

A new image denoising scheme using support vector machine classification in shiftable complex directional pyramid domain

Edge-preserving image denoising has become a very intensive research topic. In this paper, we propose a new image denoising scheme using support vector machine (SVM) classification in shiftable complex directional pyramid (PDTDFB) domain. Firstly, the noisy image is decomposed into different subbands of frequency and orientation responses using a PDTDFB transform. Secondly, the feature vector for a pixel in a noisy image is formed by the spatial regularity in PDTDFB domain, and the least squares support vector machine (LS-SVM) model is obtained by training. Then the PDTDFB detail coefficients are divided into two classes (edge-related coefficients and noise-related ones) by LS-SVM training model. Finally, the detail subbands of PDTDFB coefficients are denoised by using the different parameters to control the multiscale and multidirectional anisotropic diffusion. Extensive experimental results demonstrate that our method can obtain better performances in terms of both subjective and objective evaluations than those state-of-the-art denoising techniques. Especially, the proposed method can preserve edges very well while removing noise.