ICTACT Journal on Image and Video Processing

J. Nithya ¹, M. Madheswaran ²

Affiliations
1 Department of Information Technology, K. S. Rangasamy College of Technology, IN
2 Department of Electronics and Communication Engineering, Mahendra Engineering College, IN

Abstract

The random speckle noise in the acquired fetal ultrasound images is caused by the interference of reflected ultrasound wave fronts. The presence of speckle noise will degrade the quality of the image and even hide image details, which in turn affect the process of image segmentation, feature extraction and recognition and most importantly disease diagnosis. The standardization of measurements from the fetal ultrasound images will help the physicians to make correct diagnosis. The accuracy of diagnosis is possible only when the image is noise free. Hence it is very much important to perform filtering of the speckle noise. It is proposed that curvelet transform serves as a better edge preserving filter compared to other speckle reducing anisotropic diffusion filters. Curvelet transform is designed to handle images which involve curves using only a less number of coefficients. Hence a multiscale representation called curvelet transform is applied to enhance the visual quality of the ultrasound images. The experimented results indicate that the proposed curvelet denoising suppresses the noise effectively both in quantitative and visual means by producing high PSNR.

Keywords

Speckle Noise, Despeckling, Curvelet Transform, Anisotropic Diffusion.

Full Text

     

D. Sasikala ¹, M. Madheswaran ²

Affiliations
1 Department of Electronics and Communication Engineering, Vivekanandha College of Engineering for Women, IN
2 Mahendra Engineering College, IN

Abstract

Speckle is modeled as a signal dependent noise, which tends to reduce the image resolution and contrast, thereby reducing the diagnostic values of the ultrasound imaging modality. Reduction of speckle noise is one of the most important processes to increase the quality of biomedical images. Filters are used to improve the quality of ultrasound images by removing the noise. This paper compares the performance of the thresholding technique Bayes Shrink in despeckling the medical ultrasound images with other classical speckle reduction filters like Lee, Frost, Median, Kaun, Wavelet Bayes, Anisotropic diffusion and Wavelet. The performance of these filters is analyzed by the statistical measures such as Peak Signal-to Noise Ratio, Mean Square Error and Equivalent Number of Looks. To produce a better quality resolution picture, the filter should have high Peak Signal to Noise Ratio, low Mean Square Error, high Equivalent Number of Looks. The results obtained are presented in the form of filtered images, statistical tables and graphs. Finally, the best filter has been recommended based on the statistical and experimental results. From the results obtained Lee and Frost filter outperforms the other mentioned filters in terms of high PSNR and low MSE for high variance of noise where as anisotropic diffusion filter outperforms with high PSNR and low MSE with maximum ENL for low variance values of noise.

Keywords

Speckle Noise, Image Denoising, Wavelet Thresholding, Filters.

Full Text

     

M. Madheswaran ¹, S. Jerald Jeba Kumar ²

Affiliations
1 Center for Advanced Research, Muthayammal Engineering College, IN
2 Department of Electronics and Communication Engineering, The Rajaas Engineering College, IN

Abstract

An improved Computer Aided Clinical Decision Support System has been developed for grading the retinal images using neural network and presented in this paper. Hard exudates, Cotton wool spots, large plaque hard exudates, Microaneurysms and Hemorrhages have been extracted. SVM classifiers have been used for classification. Further rule based classifiers have been used to grade the retinal images. The percentages of sensitivity, specificity have been found for both bright lesions and dark lesions. The accuracy of the proposed method is capable of detecting the bright and dark lesions sharply with an average accuracy of 98.19% and 97.51% respectively.

Keywords

Bright Lesion, Dark Lesion, Hard Exudates, Cotton Wool Spots, LPHE, Microaneurysms, Hemorrhages, SVM, Diabetic Retinopathy (DR), Non Proliferative Diabetic Retinopathy (NPDR).

Full Text

     

M. Madheswaran ¹, S. Saravanan ¹

Affiliations
1 Centre for Advanced Research, Department of Electronics and Communication Engineering, Muthayammal Engineering College, IN

Abstract

The design of low power high performance Multiply and Accumulate (MAC) unit is presented in this paper. The power analysis for MAC unit is carried out for image filtering application exploiting insignificant bits in pixel values. The developed technique is found to reduce dynamic power consumption by analyzing the bit patterns in the input data which reduces the switching activities. The power consumption of the developed multiplier is compared with existing multiplier techniques and found that is performs better. It is observed from the simulation using SYNOPSIS EDA tool that the proposed pixel properties reusability technique saves power up to 88% with small area over head when used in MAC unit.

Keywords

Low Power, VLSI Design, Booth Multiplier, MAC.

Full Text