基于语义信息的SLAM特征提取方法研究

Research on feature extraction method combined with semantic information

视觉SLAM技术是无人驾驶技术中的关键一环,可以为无人车提供环境地图信息以及精确的位置信息。直接使用传统的特征提取算法而不加改进,不利于视觉SLAM系统的位姿解算。基于语义分割对视觉SLAM技术中特征提取的环节进行了优化,使其特征点分布更加均匀,有利于无人车进行更加准确的位姿估计,并设计实验对方法的有效性进行了验证。实验结果表明:采用改进的方法有利于视觉SLAM系统进行更准确的定位和地图构建。

Visual SLAM technology is a key part of unmanned driving technology,which can provide unmanned vehicles with environmental map information and precise location information.However,directly using the traditional feature extraction algorithm without improvement is not conducive to the pose calculation of the visual SLAM system.Based on semantic segmentation,this paper optimizes the feature extraction link in visual SLAM technology to make the distribution of feature points more uniform,and conducive to more accurate pose estimation of unmanned vehicles,and designs experiments to verify the effectiveness of the method.The experimental results show that the improved method is beneficial to the visual SLAM system for more accurate localization and map construction.