Quantum image encoding and processing : a comparative analysis of quantum computing systems
Year: 2025
Type: other
Abstract: In this work, the authors present a comparative analysis of quantum computers in the context of image processing. The benchmark was performed on seven quantum computers and one simulator. For the real quantum devices, three different families of QPU (quantum processing units) were used: ion-trap devices and superconducting qubits (the IBM Falcon and Eagle series). To encode 585 images on the quantum computers, the FRQI (Flexible Representation of Quantum Images) and LPIQE (Local Phase Image Quantum Encoding) methods were employed. The reconstructed images were compared using the MSE (Mean Squared Error) and PCC (Pearson Correlation Coefficient) metrics. The experimental protocol was implemented in Python, using the Qiskit, SciPy, and Geqie libraries. The results of the experiment are presented in tabular form, as histograms, and as a correlation matrix.This study investigated the correlation of MSE with (i) the QPU implementation technology (i.e., between ion-trap devices and superconducting qubits) and (ii) the type of superconducting quantum processor (Falcon vs. Eagle). Additionally, differences in the error distributions between these QPU series (Falcon vs. Eagle) were analyzed.The main conclusions are as follows: a strong correlation of MSE with the type of superconducting quantum processor (Falcon vs. Eagle) and a moderate correlation of MSE with QPU implementation technology (ion-trap vs. superconducting) were demonstrated. Furthermore, it was shown that Falcon-series processors exhibit a bimodal error distribution, whereas Eagle processors exhibit a unimodal distribution. No correlation was found between MSE and the error rate.
DOI: 10.34918/91245
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7125955413