FPGA Implementation of Speckle Noise Removal in Real Time Medical Images

摘要

Digital image processing aims at enhancing the potential of visual information for human interpretation. The real time Medical images are degraded by noise during acquisition and transmission process. Reducing noise in Medical images is a challenge for the researchers in Image processing field. There are different types of noises present in Digital Images. There are different types of imaging techniques used to capture the medical Images. The principle sources of noise in real time medical images are incorrect assumption of ultrasound pulse and source of reverberations. Speckle noise is the result of diffuse scattering and these images undergo constructive and destructive interferences resulting in mottled b- scan images. Speckle noise tends to reduce the resolution and contrast thereby degrades the diagnostic accuracy of this modality. Different types of filtering techniques are there in which in this paper Edge Enhanced Modified Lee filter, Modified Kuan Filter, Fast Bilateral Filter, Adaptive Centre weighted Median Exponential filter are introduced which are tested on sample ultrasound images, Magnetic resonance images. Effectiveness of these filters are compared with different types of Quality metrics using Matlab tool R2013a, in which the fast bilateral images and Adaptive Centre weighted Median Exponential filter provides the Denoise image with high quality when compared to the existing methods. The output from the Matlab is given as input to Altera Board through Quartus- II and output image is displayed via VGA monitor. By using DE1 Control panel Real time image is implemented in Altera Board.