Research On Anti-interference For Quantum Image Processing

As a rising interdisciplinary,quantum computation has become a research hotpot that many researchers focus on.In the field of cryptology,information transmission and other fields,it has achieved breakthrough progress.At the same time,people are trying to apply the advantage of quantum computation in other areas.With the development of sensor technology and liquid crystal display technology,the resolution of images that researchers capture and present are continuously improved.Traditional technology for these digital image processing has begun to appear a performance bottleneck.So people want to use the mechanics of quantum computing to solve these challenges.Quantum image processing begins to appear.However at present,this field is in the initial stage.Some algorithms such as feature extraction,image recognition are still unable to realize very well.One of the important reasons is that quantum image storage model is too ideal,which has not considered the noise in the real world.Due to the variety of noise types and intensity,there is no effective algorithm for this problem.Therefore,quantum image restoration becomes an urge problem needed to be solved.In this topic,we will propose related algorithms to deal with this problem.The main content of our study are the following aspects:1 Quantum image representation modelBefore the design of quantum image processing algorithm,we need to store the image in quantum state.Firstly,we will introduce and analyze some existing quantum image representation models,then select an optimal model NEQR to make an improvement.After model redesign,it becomes the basis of our research.2 Quantum image restoration methods with purposeIn order to design quantum image restoration method,we choose two common type of noise(Salt-and-Pepper noise and Gauss noise)to carry out targeted research of algorithm design.With the help of quantum circuit,we can verify the feasibility of our algorithms.After analysis of algorithms,the two methods we proposed can both obtain index grade performance improvement compared with the classical noise restoration algorithms.In order to better show the unique advantages of quantum algorithms,the effect of restoration is presented through simulation experiments.We also use a metric(peak signal to noise ration)to quantitatively prove the advantages of our methods.3 The research on adaptive algorithm for quantum image restorationBecause we cannot know the type of noise contained in realistic images,an adaptive algorithm for quantum image restoration is needed.In this algorithm,we will firstly judge the type of noise and then select the appropriate algorithm according to the judgment.After the design of quantum circuits,the feasibility of algorithm is verified.And with the analysis about complexity,the superior performance of quantum computation can be further proved.In the experiment,we use two common data sets to simulate the actual operation of the quantum algorithm.According to the experimental results,we can see the judgment of noise type is basically accurate and the noise processing effect is very well.Our methods not only solve the performance problems of classical image processing algorithms,but can also be used to improve the processing precision of complex algorithms such as feature extraction and pattern recognition.At the end of thesis,we will discuss the challenges and future development in the field of quantum image processing.