Ultrasonik ARK haritası işlemeye dayalı beş yöntemin karşılaştırmalı incelemesi

In this work, one newly proposed and four existing algorithms for processing ul-trasonic arc maps are compared for map-building purposes. These algorithms arethe directional maximum, Bayesian update, morphological processing, voting andthresholding, and arc-transversal median algorithm. The newly proposed method(directional maximum) has a basic consideration of the general direction of themapped surface. Through the processing of arc maps, each method aims at over-coming the intrinsic angular uncertainty of ultrasonic sensors in map building, aswell as eliminating noise and cross-talk related misreadings. The algorithms are im-plemented in computer simulations with two distinct motion-planning schemes forground coverage, wall following and Voronoi diagram tracing. As success criteria ofthe methods, mean absolute difference with the actual map/profile, fill ratio, andcomputational cost in terms of CPU time are utilized. The directional maximummethod performed superior to the existing algorithms in mean absolute error, wassatisfactory in fill ratio and performed second best in processing times. The resultsindicate various trade-offs in the choice of algorithms for arc-map processing.Keywords: Ultrasonic sensors, map building, arc maps, Bayesian update scheme,morphological processing, voting and thresholding, arc-transversal median algo-rithm, wall following, Voronoi diagram, motion planning, mobile robots.iii