Malignant skin lesions in dermoscopic images are characterized by non-uniformity of color. Such skin lesions have different shades of brown or black colors and may also have patches of pink, red, white or blue color. This fact makes segmentation of malignant skin lesions from dermoscopic images difficult than segmentation of uniform benign skin lesions. To overcome this challenge, in this paper, the hybrid pre-processing approach that uses histogram specification process is proposed. The Fuzzy C Means (FCM) segmentation method is used to demonstrate the results. This hybrid preprocessing and FCM segmentation methods are applied to segment the most common types of malignant skin lesions, Melanoma and Basal Cell Carcinoma. The results of Fuzzy C Means segmentation using different preprocessing techniques are illustrated. The results depict that the proposed hybrid approach leads to efficient segmentation of the malignant skin lesions as compared to other commonly used pre-processing approaches.