Melanoma is originated from the melanocyte producing the melanin which determines the complexion,and it has the highest mortality among skin cancers. Acral lentiginous melanoma(ALM) arises from extremities suchas hands, feet or fingernails. Since the appearance of ALM is different from melanoma on the body, conventional autodiagnosis systems for melanoma is inappropriate to detect ALM. Therefore, ALM is typically difficult to distinguishfrom general nevus, resulting in delayed diagnosis and bad prognosis. In this paper, we firstly introduce a determinationmethod for ALM by dermatologists and propose a method to rotate dermoscopic images automatically asa pre-processing for facilitating the easy determination of ALM and to select the optimal value of the Gaussian differentiationfilter parameter which is significant for precise pattern extraction using the scale space analysis. Fromexperimental results, it is shown that there exists the consistency between empirical values of the Gaussian differentialfilter parameter and optimal values derived from the scale space analysis to distinguish nevus and ALM.