Orthogonal moments based texture analysis of CT liver images

Liver tumor diagnosis by gray level and contourlet coefficients texture analysis

“…Luyao Wang et al [5] extracted texture features based on first order statistics, spatial gray level dependence matrix, gray level difference matrix and gray level run length matrix in order to classify hepatic tissues into primary hepatic carcinoma, hemangioma and normal liver with a total accuracy rate of 97.78% by an SVM classifier. Subbiah Bharathi and Ganesan [6] used Zernike features to classify normal liver with an accuracy of 98.33% and hepatocellular carcinoma with 90.67% accuracy. They also used Legendre features to classify normal liver with 97.66% accuracy and hepatocellular carcinoma with 81.67% of accuracy.…”

“…It seems that there is a lack of automatic system developed for detection and classification of liver tumors [1]- [6]. Manual segmentation is often non-systematic and time conswning.…”

Contourlet Transform based Computer-Aided Diagnosis system for liver tumors on computed tomography images

“…Luyao Wang et al [5] extracted texture features based on first order statistics, spatial gray level dependence matrix, gray level run length matrix and gray level difference matrix in order to classify hepatic tissues into primary hepatic carcinoma, hemangioma and normal liver using SVM classifier with a total accuracy rate of 97.78%. Subbiah Bharathi and Ganesan [6] used Zernike features to classify normal liver with 98.33% and hepatocellular carcinoma with 90.67% accuracy and Legendre features to classify normal liver with 97.66% accuracy and hepatocellular carcinoma with 81.67% accuracy. Mala and Sadasivam [7] characterized the liver tumor as benign or malignant using wavelet based texture analysis and Linear Vector Quantization (LVQ) neural network with an accuracy of 92%.…”

Fast and low-complexity method for exact computation of 3D Legendre moments

“…Legendre orthogonal moments can be used to represent an image with the minimum amount of information redundancy (Teh and Chin, 1988). Based on these attractive properties, Legendre moments are used in many applications such as pattern recognition (Chong et al, 2004, Luo andLin, 2007), face recognition (Haddadnia et al, 2001), line fitting (Qjidaa and Radouane, 1999), texture analysis (Bharathi and Ganesan, 2008), template matching (Omachi andOmachi, 2007, Hosny, 2010b), palm-print verification (Pang et al, 2003), occupant classification system for automotive airbag suppression (Farmer and Jain, 2003), comparison of two-dimensional polyacrylamide gel electrophoresis maps images (Marengoa et al, 2005), retrieval and classification (Yadav et al, 2008), and tool wear monitoring (Barreiro et al, 2008). It is well known that, the direct computation of Legendre moments is time consuming process and the computational complexity increased by increasing the moment order.…”