Scaling up and down of 3-D floating-point data in quantum computation

Area of Science:

• Quantum Computing
• Quantum Image Processing
• Computational Mathematics

Background:

• Quantum computation offers significant performance advantages.
• Quantum image scaling is a common geometric transformation, but lacks a floating-point data version.
• Existing methods like nearest-neighbor and bilinear interpolation are unsuitable for high-dimensional data.

Purpose of the Study:

• To develop a quantum scaling up and down scheme for 2-D and 3-D floating-point data.
• To address the absence of quantum floating-point data scaling in quantum image processing.
• To improve the precision of quantum floating-point algorithms.

Main Methods:

• Utilizing trilinear interpolation for 3-D floating-point data scaling.
• Designing a Converter module for arbitrary-size fixed-point to floating-point data conversion using [Formula: see text] qubits based on IEEE-754 format.
• Developing quantum scaling circuits for 3-D floating-point data.

Main Results:

• A novel quantum scaling scheme for 3-D floating-point data is presented.
• The proposed scheme offers improved precision for quantum floating-point algorithms compared to classical methods.
• A Converter module enables flexible data type conversion for quantum computations.

Conclusions:

• Trilinear interpolation is effective for quantum scaling of 3-D floating-point data.
• The developed quantum scaling circuits provide a practical implementation for 3-D quantum image processing.
• This work advances quantum image processing capabilities for high-dimensional floating-point data.