Quantum Chirp Transform for Image Compression and Transmission with Multi-Stage U-Net-Based Image Denoising and Reconstruction

Preserving perceptual quality and structural fidelity during image transmission remains challenging under bandwidth constraints and noisy channel conditions. Conventional compression standards often exhibit significant performance degradation under severe channel impairments, while integrated quantum-inspired compression and transmission frameworks remain largely underexplored. To address these limitations, this work proposes a simulation-based quantum-inspired image transmission framework that combines Quantum Chirp Transform (QCT)-based compression with a multi-stage U-Net reconstruction and denoising mechanism. In the proposed framework, image bitstreams are encoded using variable-dimensional representations with encoding dimension k, transformed into a chirp-structured domain, and transmitted through a numerically simulated composite quantum noise channel. The QCT exploits non-stationary quadratic phase characteristics to achieve efficient compression while preserving structurally significant image information. At the receiver, inverse processing and adaptive multi-stage U-Net enhancement are employed to suppress channel-induced distortions and improve reconstruction quality. Simulation results demonstrate compression ratios ranging from 2:1 to 128:1 depending on the selected encoding dimension, while maintaining high reconstruction fidelity. Compared with quantum Fourier transform (QFT) compression under identical transmission conditions, the proposed framework achieves superior robustness under noisy channels, with PSNR improvements of up to 4.9 dB over a QFT-based baseline and classification accuracy improvements from 84.3% to 90.4% at 10 dB SNR. Results further show that higher-dimensional encoding improves compression efficiency but increases sensitivity to channel impairments, which is effectively mitigated by the proposed multi-stage U-Net reconstruction strategy. These findings demonstrate the potential of chirp-structured quantum-inspired representations for robust image compression and transmission in bandwidth-constrained environments.