Semi-automatic optimal route location using high resolution satellite images, GIS, linear programming and active contours optimization techniques

In this paper, a novel method which has been developed for preliminary optimal general route location is presented. The developed method has two main stages. In first stage, an initial set of proposed route locations are optimized individually using active contours (Snakes) algorithm based on satellite image grey level values. Afterwards, in the second stage, the resulting alternative optimal route locations are evaluated using a decision support tool (Route_Tools) developed in MICROSOFT WINDOWS Visual C++6 to automatically select an overall optimal route location. The developed (Route_Tools) uses Linear Programming (LP) as an optimization strategy to determine an overall preliminary optimum route location using a weighted average based on three criteria, which are route total distance, route segment grades, and degree of curvature between contiguous route segments. It is being concluded that the method is feasible for rural areas where there are few objects with homogeneous radiometric characteristics.