Departmental Bulletin Paper 自律移動ロボットのための強化学習を用いたランドマーク検出手法
ジリツ イドウ ロボット ノ タメ ノ キョウカ ガクシュウ オ モチイタ ランドマーク ケンシュツ シュホウ
The Landmark Detection Method using Reinforcement Learning for an Autonomous Mobile Robot

小山, 裕昭  ,  曽我, 健太  ,  原, 翔悟  ,  元木, 誠||モトキ, マコト||Makoto, MOTOKI

59pp.47 - 57 , 2016-03 , 関東学院大学工学会
Multi-agent systems have characteristics of autonomy, cooperation and so on. In addition, they have ability to solve problems flexibly. One of the Multi-agent systems application is following behavior. In a factory, following type mobile robot using the localization method by the landmark can carry many things together without wasting time and labor. The landmark detection method using features have been proposed. However, the detection methods using features cannot detect the landmark from many features similar to the landmark in an actual environment yet. Therefore, in this study, we propose the landmark detection method that selects the optimum detector based on SURF in real time by reinforcement learning for an autonomous mobile robot in an actual environment. The result of verification experiments shows that the proposed method is the better detection precision than single detector.

Number of accesses :  

Other information