Analysis Of Feature Extraction Algorithm Using Two Dimensional Discrete Wavelet Transforms In Mammograms To Detect Microcalcifications

The paper focusing on the analysis of the feature extraction (FE) algorithm using a two-dimensional discrete wavelet transform (DWT) in mammogram to detect microcalcifications. We are extracting nine features namely Mean(M), Standard deviation(SD), Variance(V), Covariance(Co_V), Entropy(E), Energy(En), Kurtosis(K), Area(A) and Sum(S). FE is a tool for dimensionality reduction. Instead of too much information the data should be reduced representation of the input size. The extraction has been done to construct the dataset in the proper format of the segmented tumor, then only it can be given to the classifier tools and achieving high classification accuracy. The nine statistical features are extracted from the mammogram is determined the effectiveness of the proposed technique. This experiment has been conducted for 322 mammogram images and in this paper, we listed only a few. Actually, the detection of microcalcifications has been done with various algorithms discussed below. The analysis of the FE algorithm using two-dimensional discrete wavelet transform in mammogram to detect macrocalcification technique has been analyzed using MATLAB.