Cloud robot: semantic map building for intelligent service task

When a robot provides intelligent services, it needs to obtain a semantic map of the complex environment. The robot’s vision is commonly used to obtain the semantic concepts and relations of objects and rooms in indoor environments, which are labeled semantic information on the map. In an actual indoor environment, because of the great variety of objects and complex arrangements, a key problem is building a semantic map successfully in which the scale of the semantic database is large and the query speed is highly efficient. However, this is often a difficult problem to solve. Combined with cloud technology, the semantic acquisition structure of an environment based on the cloud is constructed by designing a cloud semantic database; the cloud robot can not only obtain the geometric description of the environment but also obtain the semantic map that contains the objects’ relationships based on a rich semantic database of the complex environment. It solves the problems of low-reliability when adding semantic information, errors in updating the map and the lack of scalability in the process of constructing the semantic map. An SVM (Support Vector Machine) algorithm is used to classify the semantic subdatabase on the foundation of which the feature model database is formed by extracting key feature points based on network text classification. Combining the semantic subdatabase with the semantic classification list, the objects can be quickly identified. Based on the abundant cloud semantic database, the cloud semantic map for intelligent service tasks can be implemented. The classification efficiency of the simulated experiments in the semantic database is analyzed, and the validity of the method is verified.