A Kernel SVM Classifier for Classification of Brain Tumors in Magnetic Resonance Images

Vol. 3  Issue 3
Year: 2016
Issue:Jul-Sep
Title:A Kernel SVM Classifier for Classification of Brain Tumors in Magnetic Resonance Images
Author Name:T. Chandra Sekhar Rao and G. Sreenivasulu
Synopsis:
The term Computer Aided Diagnosis (CAD) broadly encompasses the use of computer algorithms to aid in the process of image interpretation. CAD is also now used in general to categorize and computerize the extraction of quantitative measurements from medical images. CAD system has become the most important research subject in the domain of medical imaging and diagnostic radiology. CAD systems act as a credible secondary opinion thereby improving the accuracy and the consistency of radiological diagnosis. In this work a classifier based on Support Vector Machine (SVM) has been designed and presented for the classification of brain tumors in images from Magnetic Resonance Imaging (MRI). The SVM classifier uses a kernel in the form of Gaussian Radial Basis function kernel (GRB kernel) to enhance the classifier performance. The result of the classifier performance has been validated with the help of expert clinical opinion. The results demonstrate the suitability of the proposed classifier in the classification of brain tumors.

Character Restoration in Degraded Documents Using Hybrid Neuron Fuzzy Approach

Vol. 3  Issue 3
Year: 2016
Issue:Jul-Sep
Title:Character Restoration in Degraded Documents Using Hybrid Neuron Fuzzy Approach
Author Name:Harshmani, Nancy Gupta and Gurpreet Kaur
Synopsis:
Ancient documents may contain the valuable information of our historical past, which might be available in the printed or handwritten form. The text preservation of these antique documents has a vital significance for future generation and references. But due to several degradation factors such as bleeding through, shadow through, ink bleeding, paper aging, etc., documents are unable to show their contents up to the mark. There are various restoration techniques that may serve this purpose, but due to non-linear and complex nature of the degrading parameters, the existing techniques turn out to be less promising. The aim of study is to investigate the capability of ANN & Fuzzy logic, i.e. 'Neuro-Fuzzy technique' to restore the historical documents from their digital images. In the proposed technique, the Back- Propagation Neural Network (BPNN) is trained to cope with different degrading factors and Fuzzy rules are used to further suppress leftover spurious pixels. The output results of the proposed technique on different degraded document images are presented and compared with various existing techniques, viz. Otsu, Sauvola, Wolf, Niblack, Bernsen and Maximum entropy. The comparative results show the superiority of the proposed technique, which outperforms all other comparative techniques by providing visually better output images.

Breast Cancer Diagnosis from Low Intensity Asymmetry Thermogram Breast Images Using Fast Support Vector Machine

Vol. 3  Issue 3
Year: 2016
Issue:Jul-Sep
Title:Breast Cancer Diagnosis from Low Intensity Asymmetry Thermogram Breast Images Using Fast Support Vector Machine
Author Name:M. Subi Stalin and R. Kalaimagal 
Synopsis:
Breast cancer is the leading cause of death, linked to primary cancer. Screening of Thermogram images, the most robust method for early diagnosis of breast cancer is widely recommended with the introduction of several Computer Aided Diagnosis (CAD) techniques. The main difficulties of the thermography asymmetrical temperature distribution leads to abnormality for even disease. The authors have presented one of the fastest pattern recognition techniques that have been more efficient in the classification of tumors as benign or malignant – Fast Support Vector Machine (FSVM). This method has been developed and implemented in statistical learning theory over the past decade and they gave promising classification results for efficient tumor diagnosis. The main objective of the proposed work is to help in diseases diagnosis by Thermogram analyses applying a three phase approach. In the first phase of work, Thermogram images apply preprocessing and is segmented by separating left and right portions of the breast regions. After segmentation process (second phase), some textural features are extracted using Discrete Curvelet Transform (DCT): temperature range, mean temperature, standard-deviation, and the quantization of higher tone in an eight level pasteurization. This last feature considers the entire image temperature and measures the percentage of area occupied by pixels with the higher temperatures of the image. In the final phase of the work, a supervised learning method based on Fast Support Vector Machine (FSVM) is used for the extracted feature classification. The features are extracted from a set of 50 images confirmed by physician diagnosis. The proposed method achieved the average results of accuracy 98.5%, sensitivity 96%, and specificity 96.5%.

Denoising of Images by Wavelets and Contourlets Using Bi-Shrink Filter

Vol. 3  Issue 3
Year: 2016
Issue:Jul-Sep
Title:Denoising of Images by Wavelets and Contourlets Using Bi-Shrink Filter
Author Name:S. Swarnalatha and P. Satyanarayana 
Synopsis:
Denoising refers to the recovery of an image that has been contaminated by noise due to poor quality of image acquisition and transmission. Accordingly, there is a need to reduce the noise present in the image as a consequence to produce the denoised image. This paper presents Image denoising using Wavelet transforms and Contourlet transforms governed by bivariate shrinkage (Bi-shrink) filter techniques. The Wavelet transforms have the shift sensitivity and poor directionality that is shown by peak signal-to-noise ratio. In this paper, Translation Invariant Contourlet Transforms is proposed to overcome the limitations of wavelet transforms, hence to increase the peak signal-to-noise ratio. The results illustrate the efficacy of the proposed transform in terms of peak signal-to-noise ratio, execution time and visual quality of images.

Comparative Study of PCA and SPIHT Methods in Medical image Compression

Vol. 3  Issue 3
Year: 2016
Issue:Jul-Sep
Title:Comparative Study of PCA and SPIHT Methods in Medical image Compression
Author Name:Ravi Kiran, Chandrashekhar Kamargaonkar and Monisha Sharma
Synopsis:
Compression of medical picture has acquired great attention attributable to its raising need to decrease the picture size while not compromising the diagnostically crucial medical data exhibited on the picture. The PCA algorithm may be used to help in picture compression. In this paper, a comparative study is provided for PCA and SPIHT compression method. Here the PCA algorithm is characterized in two forms, i.e. Standard PCA and Block-Based PCA. The block based PCA has 2 extended-PCA algorithms that manipulate the block data of the picture are evaluated. The first algorithm is referred to as block-by-block PCA where the standard PCA algorithm is utilized on every block of the picture. In the next algorithm- the block-to-row PCA, all of block data are initially concatenated into a row before the standard PCA algorithm is therefore utilized in the remodelled matrix. In this work, the SPIHT is being compared with the above two methods in terms of image quality and compression ratio. With this work, it is observed that block-based PCA performs superior to the PCA algorithm and SPIHT with regards to picture quality, producing a similar compression ratio like the PCA algorithm.

Performance Analysis of Two Level Multiple Image Formats

Vol. 3  Issue 2
Year: 2016
Issue:Apr-Jun
Title:Performance Analysis of Two Level Multiple Image Formats
Author Name:Prakriti Gautam and Deepak Sharma
Synopsis:
Steganography is the process of hiding information in a carrier in order to provide the secrecy of text, music, audio and images. It can be defined as the study of imperceptible communication that deals with the ways of concealing the existence of communicating messages. For hiding secret information in various file formats, there exists a large variety of Steganographic techniques where some are more complex than others and all of them have respective strong and weak points. The aim of the paper is to provide the user, a comparative analysis of the first and second level steganography using the MLSB embedding technique in various types of file formats as the cover image like JPEG, BMP, PNG, etc. With the help of results obtained, the best format for both the levels of hiding can be concluded. The secret image used is also taken of all different formats so that a conclusion can be drawn that which type of format is suitable for the cover as well as secret image.

Comparative Analysis Of Edge Detection Methods And Its Implementation Using UTLP Kit

Vol. 3  Issue 2
Year: 2016
Issue:Apr-Jun
Title:Comparative Analysis Of Edge Detection Methods And Its Implementation Using UTLP Kit
Author Name:R. Hemalatha, N. Santhiyakumari, M. Madheswaran and S. Suresh
Synopsis:
In today's scenario, several advanced techniques are used in diagnosing images prevailed from medical imaging system. Edge detection methods reduce the quantity of data and remove ineffective information while preserving important structural properties in an image. The objective of this paper is to compare the performance of different edge detection methods like canny deriche, Morpho gradient, Prewitt, Ridge Valley, Roberts, Sobel and Zero are crossing of the Common Carotid Artery image. The statistical parameters such as minimum and maximum pixel values, mean, standard deviation, skewness and kurtosis are considered to study the performance of various edge detection methods with the aid of Aphelion Dev software. The edge detecting image has been implemented in Unified Technology Learning Platform to increase the processing speed of an image. It has been observed that Canny Deriche edge detection method is more suitable than other methods for detecting accurate edges. This can be used in medical applications to detect and extract the features of an image.