Suhaila Sari

Progressive Pixel-to-Pixel Evaluation to Obtain the Hard and Smooth Region for Image Compression

In this paper, we propose a new image compression scheme that is based on the wavelet transform. The image is first divided into 64 × 64 pixels block. The progressive pixel-to-pixel evaluation is done to find the individual pixel intensity in each block. Then the different value between the maximum pixel intensity value and the minimum pixel intensity value is calculated. This value will lead to a special representation of each where the block with high difference value is set as the hard region block while the block with low difference value is set as the smooth region block. This region will be decomposed individually using Discrete Wavelet Transform (DWT) and thresholded using different threshold method based on the block characteristic. The proposed method shows compression did well on image with a large smooth area.

Development of denoising method for digital image in low-light condition

A good noise reduction is a method that can reduce the noise level and preserve the details of the image. This paper proposes the development of a denoising method through hybridization of bilateral filters and wavelet thresholding for digital images in low light condition.In the first stage, the noisy image is passed through Bilateral Filter. However, only some amount of noise gets reduced and the image gives a blurred appearance. Hence to preserve the edge details and reduce the blur effect, Wavelet Thresholding is applied and it is proved that the proposed method is more successful to preserve the edges while suppressing the noise.

Optical tomography: A new filter technique for post processing image

Filtering process is important to make sure the resultant image is free from noise. This paper introduces a new filtering technique for optical tomography to reduce the noise or called a smearing effect. With this technique, it also can increase the Peak Signal to Noise Ratio (PSNR). This smearing effect is resultant from Linear Back Projection (LBP) algorithm that is applied in image reconstruction process. To solve the smearing effect, a new technique that use a threshold value is introduced which is called Averaging Grouping Colour (AGC) bar. The result shows that the AGC bar type 5 is the most suitable technique to be implemented in this system.

Analysis on MET Value Due to the Relation Between Echocardiogram (ECG) Signal and Human Physical Activities

Electrocardiogram (ECG) signal provides a lot of information about the cardiac functions through its patterns displayed. Besides to use ECG for cardiovascular diseases diagnostic function, the use of ECG has extended to wider areas such as monitoring the performance of athletes, lie detections, human authentication, stresses and emotion measurement. Moreover, from the ECG signal, the heart rate (HR), heart rate variability (HRV) and R-R intervals can be obtained and relate these physiological response value with human physical activities. Thus, the using of an electrocardiogram (ECG) device or wearable ECG monitor to diagnose the heart condition for the training performance evaluation is desirable. The result is recorded and assessed by analyzer (prototype of this project) for diagnosis. This work employs the Metabolic Equivalent Task (MET) to perform the level of physiological activities from volunteers. Any deviation from target heart rate value indicate the current physiological condition proportional to result of training level (either achieved the target already or need to add more the intensity, load, and duration of training). Thus, this work will assess and evaluate the proposed physical activity performance based on the assessment of ECG signal during exercise.

Minimizing median difference quantization error for image compression

This research emphasis on problem of quantization process in transform-based image compression. In specifying the quantizer, the size of interval is giving huge impact to the quantization performance. Generally, high quantization error will occurred if large interval is used at high difference value bin. Thus, quantizer needs to be design carefully to outfit the value as efficient as possible to reduce the quantization error. Contrary to the traditional approach, that apply uniform or non-uniform quantization step size, the proposed one utilize the high occurrence of zero coefficient by re-allocate the non-zero coefficient in a group for quantization. Then the proposed coder employ quantization error minimization mechanism by calculating the difference median quantization error at each quantization interval class. The results are then compared to the standard transform based algorithm and we found that the proposed algorithm compress the image effectively without harming the quality of the compressed image.

Image compression using a new adaptive standard deviation thresholding estimation at the wavelet details subbands

The process before quantization stage in compression process is a very crucial stage espeacially in application that require a high compression ratios. So, in this paper, we propose a new method of image compression that is based on reducing the wavelet coefficients in wavelet details subbands. It is based on the concept of local subband wavelet coefficients minimization to find the optimum threshold value for wavelet coefficients in each detail subbands. The proposed method decomposed the image into LL (low resolution approximate image), HL (intensity variation along column, horizontal edge), LH (intensity variation along row, vertical edge) and HH (intensity variation along diagonal). The coefficients in details subband retrived from the decomposition process is then manipulated in such a way that the nearly zero coefficient is discarded while the rest is remained. This process will reduce the unsignificant wavelet coefficient that leads to a great compression ratio while preserving the informative data to produce a good image quality as can be seen in the experiment done.

Hybridization Denoising Method for Digital Image in Low-Light Condition

A good noise reduction is a method that can reduce the noise level and preserve the details of the image. This paper proposes a denoising method through hybridization of bilateral filters and wavelet thresholding for digital image in low-light condition. The proposed method is experimented on selected night vision images and the performances are evaluated in terms of Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) and visual effects. Results demonstrate that the proposed denoising method has improved the PSNR and MSE of average performance of bilateral filter by 0.97 dB and 1.33 respectively and the average performance of wavelet thresholding has improved by 0.98 dB and 1.19 respectively.