How Easy is Matching 2D Line Models Using Local Search?

Local search is a well established and highly ef fective method for solving complex combinatorial optimiza tion problems Here local search is adapted to solve di cult geometric matching problems Matching is posed as the problem of nding the optimal many to many correspon dence mapping between a line segment model and image line segments Image data is assumed to be fragmented noisy and cluttered The algorithms presented have been used for robot navigation photo interpretation and scene under standing This paper explores how local search performs as model complexity increases image clutter increases and additional model instances are added to the image data Expected run times to nd optimal matches with con dence are determined for distinct problems involving models Non linear regression is used to estimate run time growth as a function of problem size Both polynomial and exponential growth models are t to the run time data For problems with random clutter the polynomialmodel ts bet ter and growth is comparable to that for tree search For problems involving symmetric models and multiple model instances where tree search is exponential the polynomial growth model is superior to the exponential growth model for one search algorithm and comparable for another