A Research On Quantum Inspired Medical Ultrasound Image Filtering

Digital image processing is becoming more and more widely used. Medical ultrasound image processing is one of the most important applications. The medical sonography is widely used in clinical diagnoisis. However, they are usually contaminated by speck noise. Medical ultrasound image filtering plays an important role in the further application of medical sonography. The concepts and theory of quantum mechanics can be combined with traditional image processing method to produce new image processing algorithms. This thesis combines the theory of quantum mechanics and quantum signal processing to propose quantum-inspired medical ultrasound image filtering method.The partial differential equation is very important in image filtering. This thesis combines the quantum mechanics and quantum signal processing to derivative the quantum-inspired diffusion coefficient and quantum-inspired anisotropic diffusion model. This method achieves good despeckling and edge details maintaining performance.Multiscale geometric analysis is an important tool for signal processing, including image processing. This thesis exploits dual tree complex wavelet transform(DT-CWT) which provides direction selectivity and complex wavelet which provides interscale correlations to propose quantum inspired adaptive threshold. This method gains good despeckling and edge details maintaining ability.Shearlet is a non classical wavelet analysis tool for the defect of the discrete wavelet processing multidimensional signal. This thesis examines directional selectivity of shearlet decomposition to propose the quantum inspired adaptive threshold. This method gains good despeckling and edge details maintaining performance.Finally, this thesis compares the three methods by simulation speckled image and real medical ultrasound image experiments. The results show that the proposed medical ultrasound image filtering speckle methods are effective and feasible. They can not only better filter the speck noise of medical ultrasound images, but also effectively maintain the edge details.