摘要
The progress of 5G communication,big data,and cloud computing technology has a great impact on autonomous ro-bots.To adapt to complex environments,autonomous robots need much information and intelligence,and thus,there are a lot of data processing problems.In this process,the main is-sues include information acquisition,multimodal representa-tion,and multi‐source information fusion.For instance,to sense surrounding pedestrians,vehicles,objects,buildings,and other objects in various natural and geographical,and illumi-nation environments,multiple heterogeneous types of sensors like camera,infrared sensor,LiDAR,and microwave radar usually need to be adopted for this complex environment.It poses challenges for information acquisition.After the acqui-sition,adaptive methods and theories of information repre-sentation need to be considered per task goals,environmental changes,and information types.On the basis of information representation,the fusion of information is an effective means to improve the gain of raw sensor information.It is bound to face the problems of fusion for detection,estimation,and decision‐making purposes,when the robots acquire multiple images,radar plots,or infrared observation information.
