Dynamic environment SLAM method based on YOLOv8 instance segmentation

• Main Authors: WANG YANKUN, ZHANG BING, WU LIGANG, YAO YURAN, WU ZIRUI, HU JIANING
• Format: Patent
• Language: Chinese; English
• Published: 01.03.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a dynamic environment SLAM (Simultaneous Localization and Mapping) method based on YOLOv8 instance segmentation, and belongs to the technical field of simultaneous localization and mapping. The method solves the problem that the existing method cannot guarantee the separation precision and real-time performance at the same time. According to the method, the YOLOv8 segmentation model is applied to the front end of visual SLAM, the YOLOv8 segmentation model is used for segmenting possible dynamic objects in the image, the SLAM algorithm is carried out after feature points of the dynamic objects are eliminated, and good positioning and mapping effects can be obtained. Moreover, the method improves the separation precision of the feature points of the dynamic object, and the real-time performance of the separation of the feature points of the dynamic object is ensured because only the dynamic object in the identification frame is segmented and the optical flow tracking method is adopted fo