Directional wavelet based features for colonic polyp classification

Texture classification is used in various pattern recognition applications that possess feature-liked appearance. This paper aims to compile the recent trends on the usage of feature extraction and classification methods used in the research of texture classification as well as the texture datasets used for the experiments. The study shows that the signal processing methods, such as Gabor filters and wavelets are gaining popularity but old methods such as GLCM are still used but are improved with new calculations or combined with other methods. For the classifiers, nearest neighbor algorithms are still fairly popular despite being simple and SVM has become a major classifier used in texture classification. For the datasets, Brodatz texture dataset is the most popularly used dataset despite it being old and with limited samples, other datasets are less used.

Recent trends in texture classification: a review

Fast communication: Statistical image modeling in the contourlet domain using contextual hidden Markov models

This paper presents a novel optimal filtering approach that can be integrated into video compression schemes to improve video compression performance. In the proposed approach the encoder computes the coefficients of a linear filter in an optimal way, so as to minimize the squared error between the original frame and the filtered reconstructed frame. The encoder then multiplexes the filter coefficients into the bit-stream for decoder access. Reconstructed image frames are filtered at the decoder using optimal filter coefficients to obtain improved image frames. The optimal filtering approach is evaluated in the form of a post-process filter as well as in the form of an in-loop filter. It is shown using H.263+ and H.264/AVC that the proposed approach improves video compression performance of standard compression schemes.

Optimal Post-Process/In-Loop Filtering for Improved Video Compression Performance

An ideal steganographic technique embeds secret information into a carrier cover object with virtually imperceptible modification of the cover object. Steganalysis is a technique to discover the presence of hidden embedded information in a given object. Each steganalysis method is composed of feature extraction and feature classification components. Using features that are more sensitive to information hiding yields higher success in steganalysis. So far, several steganalysis methods have been presented which extract some features from DCT or wavelet coefficients of images. Multi-scale and time-frequency localization of an image is offered by wavelets. However, wavelets are not effective in representing the images in different directions. Contourlet transform addresses this problem by providing two additional properties, directionality and anisotropy. The present paper offers an universal approach to steganalysis called CBS, which uses statistical moments of contourlet coefficients as features for analysis. After feature extraction, a non-linear SVM classifier is applied to classify cover and stego images. The efficiency of the proposed method is demonstrated by experimental investigations. The proposed steganalysis method is compared with two well-known steganalyzers against typical steganography methods. The results showed the superior performance of our method.

CBS: Contourlet-Based Steganalysis Method

Texture classification is a very important image analysis application. To successfully distinguish among texture categories, common features that can effectively characterize texture images are in need. The contourlet transform is a recently proposed two-dimensional technique for image analysis. It has been proved very efficient for representing images with fine geometrical structures, of which texture images are typical examples. In this paper, contourlet based feature extraction for texture classification has been investigated. A new feature design based on the contourlet spectral histogram has been successfully developed. With this feature design, satisfactory classification accuracy has been achieved for some typical sets of Brodatz textures. It has also been demonstrated that the design outperformed several other comparable schemes

Contourlet Spectral Histogram for Texture Classification

The contourlet transform is a recently developed two-dimensional transform technique. It is reported to be more effective than wavelets in representing smooth curvature details typical of natural images. To fully exploit the potential of contourlets in image processing and analysis applications, appropriate models are needed to describe statistical characteristics of images in the contourlet domain. In this paper, statistical contourlet image modeling techniques have been investigated. A contextual hidden Markov model, which was successfully applied to wavelet image denoising, has been adapted into the contourlet domain. The resulting contourlet contextual HMM has been tested in a denoising application with promising results, which verified its effectiveness in characterizing contourlet images

Contourlet Image Modeling with Contextual Hidden Markov Models

The discrete cosine transform (DCT) is widely used in image and video coding schemes. By applying the DCT to image blocks and quantizing the DCT coefficients, efficient image compression can be achieved. However, the quantization of the DCT coefficients introduces errors in the decompressed images, which appear as blocky artifacts. The blocky artifacts can be converted into “Gaussian-type” noise and removed via image denoising techniques. In this paper, an adaptive deblocking scheme has been developed to remove the blocky artifacts from images. Experimental results show that, with this adaptive approach, blocky artifacts have been successfully removed with improved peak signal-to-noise ratio (PSNR) values over the previous approach.

Zhiling Long

Papers

Contourlet Spectral Histogram for Texture Classification

Contourlet Image Modeling with Contextual Hidden Markov Models

Adaptive Deblocking of Images with DCT Compression.