Dazzle: Oversampled Image Reconstruction and Difference-imaging Photometry for the Nancy Grace Roman Space Telescope

Abstract We present algorithms and software for constructing high-precision difference images to detect and measure transients, such as microlensing events, in crowded stellar fields using the Nancy Grace Roman Space Telescope. Our method generates difference images by subtracting an oversampled reference image, with iterative masking to address outlier pixels. We also provide an analytic correction for small dither offset errors. Microlensing detection is achieved through a three-dimensional matched-filtering technique, optimized with Gaussian kernels to capture varying event durations, and verified through synthetic tests with high recovery rates. Transient photometry is performed via PSF fitting on difference images, using Nelder–Mead optimization for subpixel accuracy. The software, Dazzle, is available as an open-source Python package built on widely used libraries, offering accessible tools for the detection and characterization of transient phenomena in crowded fields.