Robot target recognition and positioning method and system based on rgb-d video

Abstract

The invention discloses an RGB-D video based robot target recognition and localization method and system. The target category is determined and accurate spatial position localization is acquired in a scene through the steps of target candidate extraction, recognition, time sequence consistency based confidence estimation, target segmentation optimization, position estimation and the like. In the invention, depth information of the scene is utilized, the spatial level perception ability of a recognition and localization algorithm is enhanced, the identity and the relevance of a target in a long time sequence target recognition and localization task are ensured while the video processing efficiency is improved through adopting key frame based long-short time time-space consistency constraints. In the localization process, collaborative target localization in a multi-information modal is realized through accurately segmenting the target in a planar space and evaluating the position consistency of the same target in a depth information space. The RGB-D video based robot target recognition and localization method and system are small in calculation amount, good in real-time performance and high in recognition and localization accuracy, and can be applied to robot tasks based on online visual information parsing and understanding technologies.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and more specifically relates to a method and system for robot target recognition and positioning based on RGB-D video. Background technique [0002] In recent years, with the rapid development of robot technology, machine vision technology for robot tasks has also received extensive attention from researchers. Among them, the recognition and precise positioning of the target is an important part of the robot vision problem and a prerequisite for performing subsequent tasks. [0003] Existing object recognition methods generally include two steps: extracting target information to be recognized as the basis for recognition and matching with the scene to be recognized. The traditional expression of the target to be recognized generally includes methods such as geometric shape, target appearance, and extraction of local features. Such methods often have shortcomings such as poor versatility...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the robot's working scene, the high computational complexity, and the large amount of calculation, there is no systematic, fast and convenient RGB-Depth video target recognition and precise positioning method.

Images