Abstract: The extraction of roads from aerial and satellite images is an important task within cartography and planning of new road networks. The recent developments of the Geographic Information System (GIS) require rapid updates of roads all over the planet. Over the past few decades, many automatic and semi-automatic methods were developed for the purpose of road extractions which vary in planning, techniques and application from one to another. The objective of this thesis is to develop and implement a road extraction system. The active contour models (Snakes) is a method that mainly used for image segmentations and edge detection. A modified method from the original approach called the balloon snakes is used in this thesis for road extraction where additional external forces are used to derive the snake along the road by taking its shape.The mathematical representation for the road extraction system is based on gradient descent method that is used to minimize the snake total energy, in addition to several types of forces that include internal, potential and (pressure) balloon forces to control the snakes' properties accurately. The proposed system is implemented using MATLAB version 7.1 (release14) and tested on a set of images representing different topographic features such as agricultural, desert, urban and city areas. The results proved that the proposed system is very efficient in rural areas. However, this efficiency decreases in the urban areas due to the existence of cars, trees and shadows of buildings that reduce the extraction efficiency. It is also shown that the main forces that affect the snakes are the potential and (pressure) balloon forces.