Nicholas H. Younan

JPEG2000 coding strategies for hyperspectral data

In using JPEG2000 for the coding of multiple-component, or multiband, images such as hyperspectral imagery, one must consider spectral decorrelation and rate allocation between image components, issues that concern the design of the JPEG2000 encoder and are, consequently, outside the scope of the JPEG2000 standard. Spectral decorrelation via a wavelet transform, as well as three alternative strategies for extending to multiple components the optimal codeblock-bitstream-truncation process widely used for spatial rate allocation in JPEG2000 coding of single-component imagery, are considered. Results indicate that the strategy of simultaneously truncating all code-block bitstreams from all codeblocks from all image components coupled with wavelet-based spectral decorrelation significantly outperforms the other techniques considered in terms of not only rate-distortion performance but also accuracy of unsupervised classification.

An image-adaptive watermark based on a redundant wavelet transform

An image-adaptive watermarking technique based upon a redundant wavelet transform is proposed. The redundant transform provides an overcomplete representation of the image which facilitates the identification of significant image features via a simple correlation operation across scales. Although the watermarking algorithm is image adaptive, it is not necessary for the original image to be available for successful detection of the watermark. The performance and robustness of the proposed technique is tested by applying common image-processing operations such as filtering, requantization, and JPEG compression. A quantitative measure is proposed to objectify performance; under this measure, the proposed technique outperforms a wavelet scheme based on the usual critically sampled DWT.

Hyperspectral image cube compression combining JPEG-2000 and spectral decorrelation

In this paper, a combined JPEG-2000 and spectral correlation method for hyperspectral image compression is presented. This compression scheme shows promising results. Also, various spectral decorrelation techniques are compared. The decorrelation using Karhunen-Loeve transform performs the best in terms of PSNR gain. But, since it is computationally expensive, there is no much gain over discrete cosine transform.

An exponentially tapered transmission line antenna

The analysis and design of an exponentially tapered transmission line antenna is presented. The exponentially tapered transmission line is designed to operate such that it has radiator characteristics at high frequency and serves as a matching section at low frequency. The NEC-2 is used to model the antenna at frequencies ranging from 500 MHz to 1 GHz to obtain the input impedance and the desired radiation pattern. >

Change Analysis for Hyperspectral Imagery

In this paper, the change vector analysis for hyperspectral imagery is investigated. Although plentiful change information is included in the change vectors with very high dimensionality, which permits the potential of finer change analysis, it is also very challenging to analyze these change vectors since any simple deterministic approaches using vector magnitudes and directions may not be feasible. In addition, the prior knowledge of ground truth is unavailable in most practical cases, where change analysis has to be completely unsupervised. Aiming at these difficulties, a simple but efficient relative radiometric normalization method is analyzed, and two automated approaches for change detection and classification using the hyperspectral change vectors after normalization are introduced. The experiment using real CASI datasets demonstrates the promising result.

Classification of Hyperspectral Data: A Comparative Study

In general, the analysis of hyperspectral remote sensing data by means of pattern recognition and/or classification is known to be data dependent. Thus, conventional methods for classifications may not be applicable due to the large amount of data collection used to characterize hyperspectral data in terms of optimality and computational time. In this paper, efficient classification methods of hyperspectral data are presented. Hyperspectral signatures are then extracted for eight different sample types (bare soil, soybean, mixed weeds, combination of soybean and weeds, sicklepod, entireleaf morning glory, pitted morning glory, and common cocklebur) and used for observing their spectral properties for detection and/or classification. The hyperspectral data that are analyzed in this paper are point source data collected from handheld spectrometers. The resulting data classification is reported accordingly and a comparative study among the various methods is made to ascertain their applicability to discriminate between the various sample types.

Watermarking of hyperspectral data

Watermarking is drawing the interest of researchers in many areas due to developments in sharing resources. Watermarked data helps protect ownership rights and provides a means of detecting illegal use. In this paper, an adaptive watermarking method based on the Redundant Discrete Wavelet Transform (RDWT) is proposed and applied to hyperspectral signatures.

Applying modular classifiers to mammographic mass classification

Classification is the last step in the computer aided diagnosis (CAD) system for determining whether a breast mass segmented from a digital mammogram is malignant or benign. Hence it is important to improve sensitivity at this stage. This work investigates the use of modular classifier (MoC) schemes, namely bagging and adaboost algorithms, for automated classification of mammographic masses. CAD systems containing a MoC are compared to CAD systems that contain traditional classifiers (TrC), for example single nearest mean or maximum likelihood classifiers. This study included 200 digitized mammograms, each manually segmented by a radiologist. In order to test the MoC and TrC approaches, conventional shape based features were extracted from the segmented masses. These features were then optimized using Fischers linear discriminant analysis (LDA). When no LDA was utilized, it was observed that MoC schemes increased the sensitivity from 74% to 83% over the TrC approaches. After performing LDA, the sensitivity increased from 83% to 88% for TrC and MoC schemes, respectively.

Tarp Filtering of Block-Transform Coefficients for Embedded Image Coding

Tarp filtering, an image coder with a simple implementation, is coupled with a block-based discrete cosine transform equipped with pre- and postfiltering. The prefilter reduces intra-block and inter-block correlation of the block-based coefficients, resulting in coefficients that are less correlated and thereby more suitable to tarp filtering. Experimental results show that the proposed coder achieves a significant improvement in rate-distortion performance as compared to the corresponding tarp coder in its original wavelet-based formulation for images with highly detailed content. A similar gain over JPEG2000 is seen for these same images, while, for images that are mostly smooth, the proposed coder performs comparably to JPEG2000

Soil texture classification using wavelet transform and maximum likelihood approach

In this paper, a wavelet-based soil texture classification system is proposed for identifying soil with different textures. The wavelet transform is used for feature extraction. The wavelet is a systematic and powerful tool for signal and image analysis due to its multiresolution characteristic. The maximum likelihood (ML) classifier is designed using a set of training samples. The ML parameter estimation method has been shown to give out optimal results. During the process of training and classification, the Fishers Linear Discrimination Analysis (FLDA) is incorporated for feature vector dimension reduction and optimization. Three different soil texture images, i.e., sand, silt, and clay are used for training and classification. Experimental results and discussion are presented.