Informatics Publishing Limited

T. R. Sivapriya ¹, A. R. Nadira Banu Kamal ², V. Thavavel ³

Affiliations
1 Department of Computer Science, Lady Doak College, Madurai, Tamil Nadu, IN
2 Department of MCA and Department of Computer Science, TBAK College for Women, Kilakarai, Ramnad District, Tamil Nadu, IN
3 Department of MCA, Karunya University, Coimbatore, Tamilnadu, IN

Abstract

Feature selection is of great importance in medical image classification especially neuroimaging classification for determining the most relevant features that will aid in accurate diagnosis of neuropsychological diseases. This paper presents a comparison of feature selection algorithms based on Support Vector Machine (SVM). To achieve robust performance and optimal selection of parameters involved in feature selection, and classification, prior knowledge is embedded to generate multiple versions of training and testing sets for parameter optimization. The integrated feature extraction and selection method is applied to a Structural Magnetic Resonance image based Alzheimer's dementia (AD) study with four different sets of non-demented and demented subjects. Cross-validation results of our study clearly indicate that the algorithm SVM-RFE trained with prior knowledge achieves 98% accuracy with Radial Basis Function (RBF) kernel and can improve performance of the classifier. This novel method of inculcating prior knowledge in SVM-RFE method which is tested in 4 different sets of datasets reveals that RBF kernel is found to outperform other kernels with a mean sensitivity of 97%, and thereby aids in quick and efficient classification of dementia.

Keywords

Support Vector Machine, Classification, Dementia, SVM-RFE.

Full Text

     

A. R. Nadira Banu Kamal ¹, P. Priyanga ¹

Affiliations
1 Department of Computer Science, Thassim Beevi Abdul Khader College for Women, IN

Abstract

The storage requirements for images can be excessive, if true color and a high-perceived image quality are desired. An RGB image may be viewed as a stack of three gray-scale images that when fed into the red, green and blue inputs of a color monitor, produce a color image on the screen. The abnormal size of many images leads to long, costly, transmission times. Hence, an iteration free fractal algorithm is proposed in this research paper to design an efficient search of the domain pools for colour image compression using Genetic Algorithm (GA). The proposed methodology reduces the coding process time and intensive computation tasks. Parameters such as image quality, compression ratio and coding time are analyzed. It is observed that the proposed method achieves excellent performance in image quality with reduction in storage space.

Keywords

Fractal Image Compression, Genetic Algorithm, Synthetic Code Book, Iteration Free, Colour Images.

Full Text

     

A. R. Nadira Banu Kamal ¹, S. Thamarai Selvi ²

Affiliations
1 Department of Computer Science, TBAK College for Women, Tamil Nadu, IN
2 Department of Computer Technology, MIT Campus, Anna University, Tamil Nadu, IN

Abstract

An enhanced iteration free fractal algorithm is proposed in this research paper to design an efficient domain pool for image compression. The proposed methodology reduces the coding process time, intensive computation tasks and also the memory requirements. The redundancies in the domain pool are reduced by the Linde Buzo Gray (LBG) Algorithm. For each range block, vector features such as mean value, edge strength, and texture strength are used to delete the irrelevant domain block. A pruning condition for terminating the searching process to find the best domain block from the domain pool is used. The codes are stored efficiently by comparing the values of the previous coded range blocks. The performance of the proposed method is compared with the existing iteration free fractal code for the benchmark images on the parameters like coding time, memory capacity and image quality. From the results of the computer simulation, the proposed method achieves excellent performance in coding time. The enhancement scheme for iteration free fractal image coding using vector quantization resulted in a reduction of 5.7 times and 11.5 times than the existing iteration free fractal code method for the single block partition of size 8x8 and 4x4 respectively on the Lena image for the codebook of size 16. The reduction in time is still higher in using code books of higher levels.

Keywords

Block Average, Domain Pool, Fractal Image Compression, Iteration Free Fractal Code, LBG Algorithm.

Full Text