Studying the effect of cold plasma on the blood using digital image processing and images texture analysis
This paper presents a method for automatic detection of liver in CT images using optimal texture features. As image contains noise so firstly, image is pre-processed with median filter. Regions of interests are chosen carefully from both liver and non-liver areas. Texture features are extracted from selected regions of interest using first order statistics and wavelet transform. Neural Network is used for classification of pixels into liver and non-liver areas. Accuracy of classification process depends on number of features extracted which should be chosen carefully. For careful selection of features, optimal features are selected from extracted features using genetic algorithm and used for final classification of image pixels. The method is tested on CT images and results obtained are presented both qualitatively and quantitatively.