This paper presents a number of problems concerning the practical (real) implementation of the techniques known as Quantum Image Processing. The most serious problem is the recovery of the outcomes after the quantum measurement, which will be demonstrated in this work that is equivalent to a noise measurement, and it is not considered in the literature on the subject. It is noteworthy that this is due to several factors: 1) a classical algorithm that uses Dirac's notation and then it is coded in MATLAB does not constitute a quantum algorithm, 2) the literature emphasizes the internal representation of the image but says nothing about the classical-to-quantum and quantum-to-classical interfaces and how these are affected by decoherence, 3) the literature does not mention how to implement in a practical way (at the laboratory) these proposals internal representations, 4) given that Quantum Image Processing works with generic qubits this requires measurements in all axes of the Bloch sphere, logically, and 5) among others. In return, the technique known as Quantum Boolean Image Processing is mentioned, which works with computational basis states (CBS), exclusively. This methodology allows us to avoid the problem of quantum measurement, which alters the results of the measured except in the case of CBS. Said so far is extended to quantum algorithms outside image processing too.
This paper presents a number of problems concerning the practical (real) implementation of the techniques known as Quantum Image Processing. The most serious problem is the recovery of the outcomes after the quantum measurement, which will be demonstrated in this work that is equivalent to a noise measurement, and it is not considered in the literature on the subject. It is noteworthy that this is due to several factors: 1) a classical algorithm that uses Dirac's notation and then it is coded in MATLAB does not constitute a quantum algorithm, 2) the literature emphasizes the internal representation of the image but says nothing about the classical-to-quantum and quantum-to-classical interfaces and how these are affected by decoherence, 3) the literature does not mention how to implement in a practical way (at the laboratory) these proposals internal representations, 4) given that Quantum Image Processing works with generic qubits this requires measurements in all axes of the Bloch sphere, logically, and 5) among others. In return, the technique known as Quantum Boolean Image Processing is mentioned, which works with computational basis states (CBS), exclusively. This methodology allows us to avoid the problem of quantum measurement, which alters the results of the measured except in the case of CBS. Said so far is extended to quantum algorithms outside image processing too.
A quantum Boolean image processing methodology is presented in this work, with special emphasis in image denoising. A new approach for internal image representation is outlined together with two new interfaces: classical-to-quantum and quantum-to-classical. The new quantum-Boolean image denoising called quantum Boolean mean filter (QBMF) works with computational basis states (CBS), exclusively. To achieve this, we first decompose the image into its three color components, i.e., red, green and blue. Then, we get the bitplanes for each color, e.g., 8 bits-per-pixel, i.e., 8 bitplanes-per-color. From now on, we will work with the bitplane corresponding to the most significant bit (MSB) of each color, exclusive manner. After a classical-to-quantum interface (which includes a classical inverter), we have a quantum Boolean version of the image within the quantum machine. This methodology allows us to avoid the problem of quantum measurement, which alters the results of the measured except in the case of CBS. Said so far is extended to quantum algorithms outside image processing too. After filtering of the inverted version of MSB (inside quantum machine) the result passes through a quantum-classical interface (which involves another classical inverter) and then proceeds to reassemble each color component and finally the ending filtered image. Finally, we discuss the more appropriate metrics for image denoising in a set of experimental results.
A quantum edge detector for image segmentation in optical environments is presented in this work. A Boolean version of the same detector is presented too. The quantum version of the new edge detector works with computational basis states, exclusively. This way, we can easily avoid the problem of quantum measurement retrieving the result of applying the new detector on the image. Besides, a new criterion and logic based on projections onto vertical axis of Bloch's Sphere exclusively are presented too. This approach will allow us: 1) a simpler development of logic quantum operations, where they will closer to those used in the classical logic operations, 2) building simple and robust classical-to-quantum and quantum-to-classical interfaces. Said so far is extended to quantum algorithms outside image processing too. In a special section on metric and simulations, a new metric based on the comparison between the classical and quantum versions algorithms for edge detection of images is presented. Notable differences between the results of classical and quantum versions of such algorithms (outside and inside of quantum computer, respectively) show the existence of implementation problems involved in the experiment, and that they have not been properly modeled for optical environments. However, although they are different, the quantum results are equally valid. The latter is clearly seen in the computer simulations
In this paper, three techniques of internal image-representation in a quantum computer are compared: Flexible Representation of Quantum Images (FRQI), Novel Enhanced Quantum Representation of digital images (NEQR), and Quantum Boolean Image Processing (QBIP). All conspicuous technical items are considered in this comparison for complete analysis: i) performance as Classical-to-Quantum (Cl2Qu) interface, ii) characteristics of the employed qubits, iii) sparsity of the used internal registers, iv) number and size of the required registers, v) quality in the outcomes recovering, vi) number of required gates and its consequent accumulated noise, vi) decoherence, and vii) fidelity. These analyses and demonstrations are automatically extended to all variants of FRQI and NEQR. This study demonstrated the practical infeasibility in the implementation of FRQI and NEQR on a physical quantum computer (QPU), while QBIP has proven to be extremely successful on a) the four main quantum simulators on the cloud, b) two QPUs, and c) optical circuits from three labs. Moreover, QBIP also demonstrated its economy regarding the required resources needed for its proper function and its great robustness (immunity to noise), among other advantages, in fact, without any exceptions.
An optimal estimator of quantum states based on a modified Kalman Filter is presented in this work. Such estimator acts after state measurement, allowing to obtain an optimal estimation of quantum state resulting in the output of any quantum image algorithm. Besides, a new criteria, logic, and arithmetic based on projections onto vertical axis of Bloch Sphere exclusively are presented too. This approach will allow us: 1) a simpler development of logic and arithmetic quantum operations, where they will closer to those used in the classical digital image processing algorithms, 2) building simple and robust classical-to-quantum and quantum-to-classical interfaces. Said so far is extended to quantum algorithms outside image processing too. In a special section on metrics and simulations, three new metrics based on the comparison between the classical and quantum versions algorithms for filtering and edge detection of images are presented. Notable differences between the results of classical and quantum versions of such algorithms (outside and inside of quantum computer, respectively) show the need for modeling state and measurement noise inside estimation scheme.
Quantum computing is currently one of the fastest emerging branches of information processing due to its theoretical computation powers exceeding conventional computers by far. However, currently well-known quantum-powered algorithms are of theoretical nature and its effect on practical problems has yet to be discovered. This work reviews current approaches in context of the applicability of quantum computing for the domain of image processing with respect to 3D computed tomography. It focuses on the encoding of n-dimensional images on a quantum computer and also the quantum measurement process, i.e., the image read-out. First results show that while the encoding is indeed highly efficient and is well-suited for representing 3D voxel data as obtained from computed tomography, the decoding to a classical representation is rather expensive. The latter part is also very sensitive to quantum hardware noise as was evidenced by both noisy simulator and real quantum hardware experiments.
Quantum image processing is paramount in image processing-based research areas nowadays due to the less computational time. This paper presents the quantum mechanic-based image denoising model. Also, This article discusses the quantum boolean mean filter and new quantum-boolean image denoising that are connected to computational basis states. This paper describes basic quantum information processing principles, new and better image representation approaches, quantum image processing implementation problems, quantum boolean mean filtering approach for image denoising observation using appropriate matrice, and finally, a work proposal for future work and conclusion. A comparison table of different classification techniques is also included in the paper.