Study On Image Operation And Watermarking Based On Quantum Image Representation

Quantum image processing,as an interdisciplinary subject of quantum information,quantum mechanics,mathematics and image processing,has excellent storage performance and ultra-parallel computing performance.Taking advantage of quantum image processing performance,this paper designs several common quantum image algorithms and quantum image watermarking algorithms for different quantum image representation models.The specific research work is summarized as follows:1.Propose common algorithms based on Generalized Quantum Image Representation(GNEQR),including quantum image binarization,quantum image background removal,quantum image local enhancement and quantum image entire enhancement,and design the corresponding quantum circuit,and the effectiveness is verified by simulation.These algorithms proposed in this paper does not need to use auxiliary qubits,thereby reducing the overall width of the algorithm implementation circuit(total number of qubits).2.Combining quantum wavelet transform and controlled-rotating gate,we propose a quantum image watermarking algorithm for the modified flexible representation of quantum images(FRQIM)and the normal arbitrary superposition state(NASS).Firstly,the wavelet coefficients of the image are obtained by quantum wavelet transform,then the rotating gate is designed by the information of the watermark image,then the embedding position of the watermark image information is controlled by the controlled gate,and finally the watermark information is embedded in the carrier image by the controlled-rotating gate.This scheme makes full use of the parallelism of quantum computing and can operate on two images at the same time.The simulation results show that for different image combinations,the embedded image is not significantly different from the original image.Compared with the quantum image watermarking algorithm that uses the quantum Fourier transform,the image distortion after watermarking is smaller,and there are more options for embedding the image.