DISCRETE RANGE CLUSTERING USING MONTE-CARLO METHODS

For automatic obstacle avoidance guidance during rotorcraft low altitu de flight a reliable model of the nearby environment is needed. Such a model may be constructed by applying surface fitting techniques to th e dense range map obtained by active sensing using radars. However, fo r covertness passive sensing techniques using electro-optic sensors is desirable. As opposed to the dense range map obtained via active sens ing, passive sensing algorithms produce reliable range at sparse locat ions and, therefore, surface fitting techniques to fill the gaps in th e range measurement are not directly applicable. Both, for automatic g uidance and as a display for aiding the pilot, these discrete ranges n eed to be grouped into sets which correspond to objects in the nearby environment. The focus of this paper is on using Monte Carlo methods f or clustering range points into meaningful groups. We compare three di fferent approaches and present results of application of these algorit hms to an image sequence acquired by onboard cameras during a helicopt er flight, Starting with an initial grouping, these algorithms are ite ratively applied with a new group creation algorithm to determine the optimal number of groups and the optimal group membership, The results indicate that the Simulated Annealing methods do not offer any signif icant advantage over the basic Monte Carlo method for this discrete op timization problem.