Segmentatation based feature extraction of MRI images using Wavelet and implementation on FPGA

This paper presents a hardware implementation approach for the feature extraction of MRI images using wavelet transform. In this paper wavelet based texture feature extraction method has been proposed. Which has mainly two steps (i) The input MRI image is decomposed into different frequency sub band images using 2-D Discrete Wavelet Transform, (ii) by using Segmentation based Fractal texture analysis (SFTA) the extraction of texture features are done from the decomposed Low frequency image. For evaluating the Performance of this proposed technique the comparison with GLCM and Haralick's texture feature are done in terms of MSE and classification accuracy. It also gives better accuracy and reduced Mean squared error (MSE). Thus, the resultant WSFTA technique found to be more useful and performs more accurately than the previous experiments Field programmable gate array (FPGA) implementation of proposed work provides benefit like parallel processing and runtime re-configurability. The FPGA implementation is done using Xilinx Vivado environment and Zedboard which is optimized for image and video processing IPs.