| 摘要: |
| 本文提出了一种基于激光雷达和深度相机数据的区域划分方法，通过平均距离检测、宽度判断和平滑性分析实现区域入口的高效识别；结合目标检测技术和深度信息，精确计算物体的三维空间位置，并通过物体类别统计划分功能性区域.本方法融合了几何信息与语义信息，不仅提升了区域入口检测的精度与效率，还在动态场景中实现了物体位置的判断和功能性区域划分.实验结果表明，本方法在区域入口检测、物体识别及区域划分的精度和实时性方面均优于传统方法，能够生成高效的语义分割区域地图. |
| 关键词: 图像处理 目标检测 语义地图 激光雷达 区域分割 |
| DOI：10.20192/j.cnki.JSHNU（NS）.2025.02.017 |
| 分类号:TP18 |
| 基金项目:上海师范大学一般科研项目（SK202123） |
|
| Real-time determination of home scenarios based on YOLOv8 object detection |
|
WANG Ruiqing, AN Kang, ZHANG Hui
|
|
College of Information, Mechanical and Electrical Engineering, Shanghai Normal University, Shanghai 201418, China
|
| Abstract: |
| A region entrance detection method based on lidar and depth camera data was proposed， by which efficient recognition of region entrance was achieved through average distance detection， width evaluation and smoothness analysis. By integrating object detection technique and depth information， the 3D spatial positions of objects were accurately calculated and functional regions were classified through object category statistics. Geometric information was integrated with semantic information in this approach， improving the accuracy and efficiency of region entrance detection while enabling real-time and robust object position estimation and functional region segmentation in dynamic scenarios. Experimental results demonstrated that the proposed method was superior to the traditional approaches in terms of precision and real-time performance for region entrance detection， object recognition and region segmentation， enabling the generation of efficient semantic region maps. |
| Key words: image processing object detection semantic map lidar region segmentation |
