The image identification methodology associates an image with a unique identifiable representation. Whenever the methodology regenerates an identifier for the same image, moreover, the newly created identifier needs to be consistent in terms of representation value. In this paper, we discuss a methodology for image identifier generation utilizing luminance correlation. We furthermore propose a method for performance enhancement of the image identifier generation. We also demonstrate the experimental evaluations for uniqueness and similarity analysis and performance improvement that have shown favorable results.