A more precise method of sky subtraction in SDSS astronomical image processing

We review a more precise method of sky subtraction for the astronomical images of the Sloan Digital Sky Survey (SDSS), which was presented by our group before. The SDSS images are originally processed by the photometric pipeline (PHOTO), which often over-estimates the sky background in processing images of galaxies with large size and/or in crowded fields. We use the second-order Legendre polynomials to fit both rows and columns of a image respectively to construct a more correct sky background model. Our method makes the subtracted-background counts follow a Gaussian distribution with a mean close to zero, which shows our sky background model is successful.