METHOD AND APPARATUS OF VISUAL FEATURE AUGMENTATION FOR VISUAL SLAM USING OBJECT LABELS

摘要

The method and device for enhancing image feature points in Visual SLAM using object labels is presented, and the method and apparatus for enhancing image feature points in Visual SLAM using object labels extracts feature points from the frames constituting the image acquired through the camera, , A visual feature processing unit that generates a visual word vector recording the distribution of descriptors including information on the characteristics of the extracted feature points, recognizes objects located in the feature point regions from the frame, and extracts semantic information about the recognized objects It may include a semantic feature processing unit and a location estimator for identifying whether the current location is a previously visited location based on the visual semantic word vector to which the semantic information is added to the visual word vector.