Dataset and reconstruction pipeline for Hadamard‑based single‑pixel imaging

This repository contains the full dataset, reconstruction notebook, and reference sample image used for the Hadamard‑based single‑pixel imaging experiments presented in this work. The dataset includes raw complementary pattern measurements acquired using a 64×64 Hadamard basis, along with a high‑resolution image of the positive USAF 1951 resolution test target used as the sample during acquisition. The Jupyter notebook provides a complete, reproducible pipeline for computing Hadamard coefficients, performing image reconstruction, and visualizing both individual and averaged measurements. All processing steps are implemented using Python, NumPy, SciPy, and Matplotlib, and the workflow is designed to be fully transparent and easy to adapt for related computational imaging experiments. Contents Jupyter Notebook: Full reconstruction workflow, including coefficient extraction, Hadamard transform, and visualization tools. Dataset: Raw intensity measurements for all complementary Hadamard patterns across multiple acquisitions. High‑resolution sample image: Ground‑truth reference image of the USAF resolution target used during acquisition. Purpose This repository enables full reproducibility of the reconstruction results, facilitates benchmarking of computational imaging algorithms, and provides a clean example dataset for researchers working on single‑pixel imaging, compressive sensing, and Hadamard‑based optical systems.