User Guide

Hierarchical Image Encoding

MHRQI uses a hierarchical qudit representation (Multi-scale Hierarchical Representation) to encode image intensity data.

The MHRQI Algorithm

1. Subdivision: The image is recursively subdivided into $d \times d$ quadrants.
2. Hierarchy: Each level of subdivision corresponds to a pair of qudits ($q_y, q_x$).
3. Encoding: Pixel intensities are mapped to rotation angles or basis states within the quantum circuit.

Quantum Denoising

The denoising component utilizes hierarchical consistency checks. If a pixel's intensity differs significantly from its parent region's average (computed in superposition), it is marked on an outcome qubit.

Usage

mhrqi --denoise --shots 100000 --size 16