融合人运动模式分析的服务机器人和谐导航

Harmonious navigation mechanisms utilizing motion patterns of people

提出一种学习人典型运动模式的方法,并利用该方法对环境中的人的行为做出预测,协调机器人以达到与人和谐共处的导航目的.算法首先通过非重叠多摄像头采集人在环境中不同地点间的运动轨迹;其次,应用两层模糊K均值算法分别对这些运动轨迹进行空间和时间序列上的分类,并利用TSC标准对每一次分类结果进行评估;然后建立每一聚类运动模式的概率方程,依此实现对摄像头网络观测下人运动行为的预测,进而调整机器人的导航策略以达到与人和谐共处的导航目的.实验展示了该算法能够快速地利用人的运动调整其导航行为.

A technique for learning collections of trajectories that characterize representative motion patterns of persons is proposed, and this algorithm is applied to predict the motions of persons and direct the robot to harmoniously coexist with people. Data recorded with a camera network is clus- tered hierarchically using fuzzy K-means algorithm based on spatial and temporal information respectively, and the quality of each clustering results is evaluated by the tightness and separation criterion （TSC）. Then each motion pattern is represented with a chain of Gaussian distributions. As a result, whenever the camera network detects a person it computes a probabilistic estimate about which motion pattern the person might be engaged in. During path planning the robot then uses this belief to improve its navigation behavior. Practical experiments demonstrate that our approach allows a robot to quickly adapt its navigation plans according to the activities of the persons.