Autoimmune disease is an immune system disorder caused due to the reaction of white blood corpuscles against the body's own cells and tissues. Indirect immunofluorescence(IIF) along with human epithelial type-2 (HEp-2) cells is a widely accepted screening technique for diagnosing autoimmune diseases. Manual interpretation of HEp-2 cell images causes inter and intra observer variations due to its subjective analysis. In addition, the presence of noise and uneven illumination in the image may lead to incorrect diagnosis. Hence, it is important to perform noise removal and illumination correction to enhance the diagnostic relevance of IIF. In this work, an attempt has been made to perform uneven illumination correction in HEp-2 images using two methods, namely, non-local Retinex (NLR) and morphological based top-hat transform (THT). Twenty eight images from a public domain database are considered in this study. Firstly, the images are filtered using anisotropic diffusion filter to perform noise removal. In the next step, the non-uniform illumination correction is performed using NLR and THT methods by selecting value channel in hue, saturation and value (HSV) color space. The performance of the methods are analyzed using edge based features. The results show that the mean edge content based contrast (MEC) is found to be 0.14 for raw image and 5.6 and 3.04 for NLR and THT respectively. This informs that NLR performs better than THT in non-uniform illumination correction of HEp-2 images. Thus, this study seems to be clinically useful. Copyright 2016, ISA All Rights Reserved.