The objective of this project was to work on the automation of both the process of quantification and the differentiation of visceral and subcutaneous adipose tissue. As a result a software prototype was created. The image acquisition is based in a single-slice protocol, taken at the inter-vertebra space L4-L5. The algorithms recognize the edges within the images to differentiate visceral and subcutaneous adipose tissue. The software gives reproducibility to the calculated areas in square centimeters, which overcomes the error introduced by the human operator and also decrease the time needed for diagnosis. Abnormal quantities of abdominal adipose tissue have been associated with metabolic disorders, hypertension, type 2 diabetes and increase in the dead rate by cardiovascular diseases. However, the computational tools available to measure fat in CT images require meticulous manipulation by users. Conventional algorithms are based only in mathematical operations and overlook some physiological information for recognizing anatomical structures. This work overcomes those problems in a systematic way providing a semi-automated software based on image processing algorithms and physiological criterion