Somkait Udomhunsakul

Feature extraction in medical MRI images

A feature extraction approach in medical magnetic resonance imaging (MRI) is proposed. In this approach, first the combination of spatial filters using the 5/spl times/5 Wiener filter followed by a 3/spl times/3 Gaussian filter is used to remove the noisy pixels while preserving the important information. Next, the edge detection algorithm based on multiple-scales edge detection of Gabor filters is applied. Moreover, wavelet transform based image fusion is used to combine the detail coefficients. Finally, the nonmaxima suppression technique is adopted to get the final result. The experiments show that the proposed algorithm can be detected well-localized, and thin edges. Therefore, the algorithm leads to a useful method for feature extraction in MRI images and can be used for diagnostic purposes.

Multi-focus Image Fusion Based on Stationary Wavelet Transform and Extended Spatial Frequency Measurement

In this paper, we propose a multi-focus image fusion approach based on Stationary Wavelet Transform (SWT) and extended the Spatial Frequency Measurements (SFM). Our proposed approach, two fused images are firstly decomposed into four subbands, which are one approximation subband (LL) and three details subbands (HL, LH and HH). Next, each subband is partitioned into blocks and each block is identified the clearer regions by computing the focus measure using the extended Spatial Frequency Measurement (SFM). Finally, the recovered fused image is reconstructed by performing the Inverse Stationary Wavelet Transform. From the experimental results, we found that the proposed method outperforms the traditional Wavelet Transform and SFM based methods in terms of objective and subjective assessments.

Wavelet filters comparison for ultrasonic image compression

At present, wavelet transform has become a standard tool in medical image compression. Although, the choice of wavelet filters is an essential issue that can be determined the compression image quality. The goal of this paper is to provide the comparison of applying different types of wavelet filters belonging to orthogonal and biorthogonal wavelets with different orders on the ultrasonic images. The wavelet filters used are Haar, Daubichies8, 6/10, 9/7 and 5/3 filters. The performance evaluation of the image quality is measured objectively using maximum difference (MD), normalized absolute error (NAE), and normalized mean square error (NMSE). Moreover, ultrasonographers do the subjective performance assessment. From the experimental evaluation, it can be concluded that objective assessment alone is not suitable objective scale to evaluate the quality of the compressed images. Therefore, subjective assessment is very important to take into account visual quality.

Image Quality Assessment for JPEG and JPEG2000

Assessment of the compressed image quality is an important issue for image processing system. In this paper, we propose a novel objective assessment to measure the quality of gray scale compressed image, which is developed from the fundamental objective measurement. It is also correlated well with human response and least time taken comparable to some existing measurements. The proposed measurement, spatial frequency measurement (SFM) is adopted to enhance the image quality evaluation performance. From the experimental results, we found that mean average error (MAE) with SFM (MAESFM) is the suitable measurements that can be used to measure the quality of JPEG and JPEG2000 compressed images. In addition, MAESFM measurement is scaled to make them equivalent to MOS, given the rate of compressed image quality from 1 to 5 (unacceptable to excellent quality).

Ultrasonic speckle denoising using the combination of wavelet transform and Wiener filter

Ultrasonic images are inherently affected by multiplicative noise, which is due to the coherent wave interference in tissue. This paper presents a method for ultrasonic speckle denoising using the combination of wavelet transform and wiener filter to effectively reduce the speckle noise while preserving the resolvable details. In our method, the steps involved are finding the 2D discrete wavelet transform of the logarithmic image. Then, the wiener filter is used to apply over areas in each detail subband (HH,HL and LH). Next the inverse wavelet transform is computed and applying the inverse logarithm. To evaluate the denoising performance, mean square error (MSE), signal to mean square error (S/mse) and edge preservation (/spl beta/) are used. From the experimental results, we found that our approach leads to an effective method for ultrasonic speckle denoising.

Comparative efficiency of Wavelet filters for multi-focus color image fusion

The comparative efficiency of Wavelet filters for multi-focus color image fusion is presented in this paper. Our experiment purposes are study and examine the suitable mother wavelets for using in multi-focus color image fusion. In fusion process, we use HSI color model and the fusing technique based on Stationary Wavelet Transform with Extended Spatial Frequency Measurement method. In our experiments, we investigate the effect of applying different types of wavelet filters belonging to orthogonal and biorthogonal wavelets with different orders. The wavelet filter used are Daubichies4, 8, 10, 12, 14, 16, 18 and Biorthogonal2.2, 2.4, 2.6, 2.8, 3.1, 3.3, 3.5, 3.7, 3.9, 4.4, 5.5, and 6.8. The fused image quality assessment is measured using Peak Signal to Noise Ratio. The result indicated that Biorthogonal 3.3 and 3.5 are proper choices for multi-focus color image fusion.

Syllable segmentation of thai human speech using stationary wavelet transform

This paper presents the Thai syllable segmentation of human speech algorithm using the 1-D stationary wavelet transform (SWT). The detail coefficient component of SWT, which represents the high frequency part of input signal, is used to analyze the boundaries of syllables. Our proposed approach, the input signals are separated into two parts. First part, the data of signal is used to find the boundaries of syllables. In the second part, the SWT is applied to the input signal. The detail coefficients of signal are used to find the center of syllables. Finally, both results are compared for syllable segmentation process. The experimental results show that the syllable segmentation of human speech using high frequency of SWT yield promising results.

Influence of SFM to objective assessment of compressed image

Evaluation of the quality of image compression still remains an important issue. In this paper, we propose a novel objective assessment to evaluate the quality of gray scale compressed images. An image characteristic measurement, Spatial Frequency Measurement (SFM), is employed to get the new objective measurements, which are Mean Square Error with SFM (MSESFM), Edge Measurement with SFM (ESFM) and Correlation Measurement with SFM (CSFM). From the experiments, we found that SFM influences to the objective assessment. In addition, the reliabilities of the new measures have been improved and better match the subjective assessment than using the traditionally simple objective assessments.

Distribution Model between Objective Measurement and Subjective Measurement

Evaluation of the quality of image compression still remains an important issue. In this work, we propose the new objective measurements, which are developed from five traditionally simple measurements (Mean Square Error, Edge Measurement, Correlation Measurement, Visual Human System and Spectral Measurement). In our study, we evaluated the quality of the compressed gray-scale images using both objective and subjective tests. We found the relationship between objective and subjective measurements from the distribution data that can be modeled and defined by exponential least square method. From the experimental results, the reliabilities (Correlation coefficient) of our new proposed measurements are better than five traditionally simple measurements.

Edge detection in ultrasonic images using Gabor filters

An integration method of speckle suppression and edge detection in ultrasonic images using Gabor filters is presented. In this approach, the logarithm is firstly applied to the ultrasonic images in order to convert the multiplicative noise into the additive noise case. The wavelet based denoising is adopted in order to remove the speckle noise in ultrasonic images. Next, the inverse wavelet transform is computed and applied the inverse logarithm. In the edge detection process, the wavelet transform using Gabor filters is applied using the third and fourth scales. The wavelet transform coefficients of the edges candidate are combined using the alternative to product with shift method where the shifting edge locations through multiple scales are considered for robust edge detection in the presence of noise. Finally, the nonmaxima suppression is used to get the final result. The result reveals that the proposed method leads to a useful method for edge detection in ultrasonic images. Also, the method is general and can be applied to other images.