In order to understand better the image description parameters generated by the FOCAS (Faint Object Classification and Analysis System) automatic astronomical image cataloging system, we have synthesized and analyzed images of dim galaxies. These simulated galaxy images were created from brighter galaxy images by geometric size reduction and intensity scaling followed by a convolution with a stellar image. The synthesized images are added at random locations to existing, digital images and the modified images are processed. The simulated galaxies are distinguished from the intrinsic images in the field by matching positions. The image parameters thus obtained serve to increase confidence in the automatic cataloging and calibrate various data reduction procedures.