Image Processing based Linear Discriminant and Quadratic Discriminant Classifier for Feature Extraction Models

An automatic method for classification of RBCs is discussed in this paper to overcome the limitations of manual analysis. This is helpful for fast, precise and efficient analysis of blood cells in order to diagnose diseases. Wavelet transform and thresholding techniques are used for feature extraction. Two classifiers and three combining rules are implemented with PR Toolbox. LDC and QDC classifiers are used separately and then combined. Highest accuracy of 85% is obtained by LDC. Meanc Productc combiner rule gives highest accuracy of 90%. Source coding is performed in MATLAB R2016a.