HARDWARE REALIZATION OF ROI BASED IMAGE COMPRESSION USING THE DWT APPROACH ON FPGA FOR MEDICAL APPLICATIONS

Image compression reduces the transmission and storage cost, which offers higher visual perception for users. The proposed ROI based image compression is designed and implemented on Artix-7 FPGA. The normal and abnormal Magnetic resonance Imaging (MRI) image databases are used for image compression with selective Region of interest (ROI). Image compression is modeled using a discrete wavelet transformation (DWT) based lifting scheme. The DWT algorithm is designed using Arthematic operations like adders, multipliers along with registers, which offers lowcost implementation, improves the hardware complexity and system performance. The ROI based image compression is synthesized on the Xilinx ISE environment and analyzes resource utilization. The proposed method utilizes 1% slices and consumes 0.108W total power. The reconstructed ROI images are analyzed using PSNR and MSE ratio. The recovered ROI based normal MRI brain image provides an average of 49.09 dB PSNR and MSE ratio of 0.827. Similarly, an average of 43.95 PSNR value and 2.758 MSE ratio for recovered ROI based on abnormal MRI brain images.