Concept learning-based thorough perception and dynamic understanding method

Abstract

According to the concept learning-based thorough perception and dynamic understanding method provided by the invention, scene analysis and information mining based on deep learning are leading fields of information science at present, are successfully applied to segmentation, classification and identification of various scenes, and obtain a large number of excellent results; on the basis of feature extraction, static reasoning based on the Bayesian network is already applied to image diagnosis or understanding, and feasibility is provided for automatically extracting scene multi-source multi-mode features and performing reasoning under a concept framework. On the basis, based on robot collaborative dynamic reasoning and perception coupling, the dynamic perception and deep understanding problems of a large-space complex scene can be effectively solved, and the accuracy is far higher than the performance of using a single deep neural network and static reasoning.

Description

technical field [0001] The invention belongs to the technical field of image understanding, in particular to a thorough perception and dynamic understanding method based on concept learning. Background technique [0002] All over the world, sudden disasters such as earthquakes, typhoons, tsunamis, and terrorist activities occur frequently, and swarm robots have major advantages in solving emergency search and rescue problems. [0003] The technical bottleneck of robot application is how to complete the environment perception, in-depth understanding and precise control under unmanned control for large-scale complex dynamic scenes. [0004] At present, there are two directions of research on environmental understanding at home and abroad: one is 3D panoramic scene understanding, which uses a single frame point cloud as input and combines small-scale context template relationships to complete the understanding of small scenes. However, in the face of panoramic scene understand...

AI Technical Summary

Problems solved by technology

However, it is impossible to establish a deep understanding relationship between objects, behaviors, and events in a large-scale complex dynamic scene, and it is impossible to realize intelligent unmanned control in the true sense.

Images