A Novel Wavelet Fractal Feature Extraction Method For Mammogram Image Classification

Abstract

Breast cancer is the severe health issue among women’s and it caused not only by women, but also by men because of the cause of smoke produced by males from the plastic industry. The challenging task of breast cancer that is, early detection for increasing the patient's survival rate. This paper introduces a novel fractal features based on Wavelet Feature extraction technique for mammogram classification. Wavelet-based fractal characteristics consist of two stages, Initially, using Discrete Wavelet Transform, the input image is decomposed into different frequency sub-band images, after which the texture characteristics are extracted from the decomposed image using Fractal Texture Segmentation Analysis. Generally, the proposed method is used to separate certain features within an image, according to a given shape. Here, the proposed approach is used to detect mammogram image features and is graded using neural network back propagation. The proposed feature extraction method Compared to existing methods such as GLCM, SFTA, WSFTA, LBP. This experiment is performed and tested on MIAS dataset to test the performance of the proposed method findings are very positive in terms of classification accuracy.