The Research Of Quantum Image Processing Algorithms And The Relative Circuits Designing

As quantum computation has the unique advantages of parallel computing,entanglement,superposition state,researchers propose many quantum versions of classical algorithms,and prove the high efficiency of the proposed algorithms.Quantum image processing is the combination of quantum computation and image processing,it has been a hot issue in present research.Quantum image processing includes geometric transformation,color transformation,image scaling,image scrambling,image segmentation,feature extraction,quantum image watermark and quantum image encryption,as well as the foundation work,quantum image representation method.This thesis focus on quantum image processing algorithms and the corresponding circuits.The main work is shown as follows:(1)Quantum multidimensional color image representation methodsQuantum image representation methods are the foundation and precondition for quantum image processing.The flexible representation of quantum images(FRQI)and the novel enhanced quantum representation(NEQR)as two frequently-used quantum image representation method,have respective advantages and disadvantages,and complement each other.However,they can only be applied to bidimensional gray-scale image.Therefore,the thesis improves FRQI and NEQR,and extends them to quantum multidimensional color image representation methods,namely EFRQI and ENEQR.(2)The research of quantum multidimensional color image scalingAs the current research of quantum image scaling is limited to bidimensional gray-scale image,thus the thesis studies quantum multidimensional color image scaling,including scaling up and scaling down.The proposed quantum multidimensional color image representation methods,EFRQI and ENEQR,are used.Image scaling is inevitable to use interpolation,but the nearest-neighbor interpolation which will be used in the paper,is only be applied to the classical bidimensional gray-scale image.Hence,the nearest-neighbor interpolation is extended to quantum multidimensional color image.Also,the circuits are designed for quantum multidimensional color image scaling.In order to optimized circuits,circularly translation(CTx)operation is proposed.(3)Global and Local translation designs of quantum image based on FRQIThis thesis designs two kinds of quantum image translation based on FRQI,including global translation and local translation.Firstly,global translation is realized by employing adder modulo N,where all pixels in the image will be moved,and the circuit of right translation is designed.Meanwhile,left translation can also be implemented by using right translation.Complexity analysis shows that the circuits of global translation in this thesis havelower complexity and cost less qubits.Secondly,local translation,consisted of one-column translation,multiple-columns translation and translation in the restricted area,is designed by adopting Gray code.In local translation,any parts of pixels in the image can be translated while other pixels remain unchanged.In order to lower complexity when the number of columns needing to be translated are more than one,multiple-columns translation is proposed,which has the approximate complexity with one-column translation.To perform multiple-columns translation,three conditions must be satisfied.In addition,all translations in this paper are cyclic.(4)Quantum Gray-scale Image Dilation and Erosion AlgorithmsGray-scale image morphological processing is extensively applied to the field of boundary detection,image segmentation and feature extraction,etc in traditional computer.Its time complexity,however,is very high.On account of the unique superiority of quantum computation,this thesis proposes a novel dilation and erosion processing algorithm based on quantum loading scheme and quantum reversible adder/subtraction circuits.Compared to electronic computer,the algorithm can store images with less qubits,and the complexity is lower.Besides,the proposed meet the optimal circuit designs.