Coadding

Coadding, also known as image stacking or median combining, is a technique used in image processing, particularly in astronomy and digital photography, to improve the signal-to-noise ratio of an image. It involves combining multiple individual images of the same scene or object into a single, composite image. This is achieved by aligning the individual images and then averaging the pixel values at each corresponding location.

The fundamental principle behind coadding is that random noise, which affects each image independently, tends to cancel itself out when averaged over multiple exposures. Signal, however, which is consistent across all images, is reinforced. Therefore, as more images are coadded, the signal-to-noise ratio of the resulting image increases, leading to a clearer and more detailed final product. This is especially crucial in astronomical observations where faint celestial objects are often difficult to discern due to their low intrinsic brightness and the presence of background noise from the atmosphere and detectors.

In practice, coadding requires precise alignment of the input images. Misalignments can introduce artifacts and blur the final image. Techniques like image registration are used to correct for shifts, rotations, and scaling differences between the frames. While simple averaging is common, other methods like median combining are also employed. Median combining can be more robust to outliers, such as cosmic ray hits or satellite trails, as it uses the median value of pixels rather than the mean, effectively ignoring extreme values. The overall outcome of coadding is a significantly improved image quality, making faint details more visible and reducing the impact of random noise.