网站访问量
今日访问量: 0
动态场景下基于激光雷达的SLAM算法
LiDAR-based SLAM Algorithm for Dynamic Environments
冯迎宾;李亚玮;王天龙;曾城锋; FENG Yingbin;LI Yawei;WANG Tianlong;ZENG Chengfeng;Shenyang Ligong University;使用激光雷达在动态场景下实现精确的位姿估计与地图映射是同时定位与建图(simultaneous localization and mapping, SLAM)研究领域的重要内容之一，但动态环境中物体移动会导致SLAM算法精度下降，为此提出一种低成本且可有效剔除动态影响的激光雷达SLAM算法。首先引入深度图投影，通过检测相邻时刻深度图之间的像素值波动，筛选并去除动态点云；然后进行地面点云分割，利用非地面点云的特征实现位姿估计和地图映射，利用地面点云的特征施加地面约束，限制高度漂移；最后引入回环检测矫正全局姿态。实验结果表明，与LOAM、LeGO-LOAM和SuMa算法相比，本文算法可更有效剔除动态目标，提供更优秀的定位建图效果和鲁棒性能。
To achieve LiDAR-based precise pose estimation and mapping in dynamic environments is a big challenge in the field of simultaneous localization and mapping(SLAM).Dynamic objects can significantly interfere with a robot's environmental perception and autonomous navigation.To address the decline in SLAM accuracy caused by moving objects in dynamic environments, this study proposes a low-cost yet effective LiDAR SLAM algorithm to eliminate dynamic interference.The algorithm introduces depth map projection to identify and remove dynamic point clouds by detecting pixel value fluctuations between consecutive depth maps.Ground point cloud segmentation is then performed, leveraging non-ground point cloud features for pose estimation and mapping.Ground point cloud features are utilized to impose ground constraints, mitigating height drift.Additionally, loop closure detection is incorporated to correct global pose drift.Experimental results demonstrate that, compared to LOAM,LeGO-LOAM,and SuMa, the proposed algorithm more effectively filters out dynamic objects, delivering superior localization and mapping performance with enhanced robustness.
基于改进YOLOv8的脑肿瘤检测模型
Brain Tumor Detection Model Based on Improved YOLOv8
张傲;刘微;刘阳;杨思瑶;管勇;李波;刘芳菲; ZHANG Ao;LIU Wei;LIU Yang;YANG Siyao;GUAN Yong;LI Bo;LIU Fangfei;为应对磁共振成像(MRI)中形态复杂和边界不规则的脑肿瘤检测，提出一种改进的YOLOv8模型TumorNet-YOLO。该模型通过三项创新模块提升检测性能：自适应感受野卷积模块增强多尺度肿瘤特征的提取能力，降低漏检率；分割融合卷积模块通过多尺度特征融合，增强浅层和深层特征的协同作用；可变形融合模块优化对不规则肿瘤区域的检测，提升模型在复杂MRI背景下的鲁棒性。实验结果表明，TumorNet-YOLO在脑肿瘤检测数据集Br35H中表现优异，平均精度均值mAP@0.5为96.6%,mAP@0.5∶0.95为73.8%。此外，模型计算量(GFLOPs)为8.6,显著优于现有方法。为了验证模型的泛化能力，在BCCD和BTOD数据集上进行了对比实验，结果显示TumorNet-YOLO在mAP@0.5和mAP@0.5∶0.95等多个指标上超越了YOLOv8n,表明TumorNet-YOLO可为脑肿瘤检测和医学图像分析提供更为有效的解决方案。
In order to address the challenges of detecting brain tumors with complex shapes and irregular boundaries in Magnetic Resonance Imaging(MRI),an improved YOLOv8 model, TumorNet-YOLO,is proposed.This model enhances detection performance through three innovative modules: the Adaptive Receptive Field Convolution Module improves the ability to extract multi-scale tumor features and reduce false negatives; the Segmentation Fusion Convolution Module strengthens the synergy between shallow and deep features through multi-scale feature fusion; the Deformable Fusion Module optimizes the detection of irregular tumor regions and improves the model's robustness in complex MRI backgrounds.Experimental results show that TumorNet-YOLO performs excellently on the brain tumor detection dataset Br35H,with a mean Average Precision mAP@0.5 of 96.6% and mAP@0.5∶0.95 of 73.8%.Furthermore, the model's computational cost is 8.6 GFLOPs, which significantly outperforms existing methods.To validate the model's generalization ability, comparative experiments were conducted on the BCCD and BTOD datasets.The results demonstrate that TumorNet-YOLO outperforms YOLOv8n in multiple metrics, including mAP@0.5 and mAP@0.5∶0.95,indicating that TumorNet-YOLO provides an effective solution for brain tumor detection and medical image analysis.
公告栏
期刊信息
期刊名称： 沈阳理工大学学报（Journal of Shenyang Ligong University）
创办日期： 1982年
主管单位： 辽宁省教育厅
主办单位： 沈阳理工大学
出版单位：《沈阳理工大学学报》编辑部
刊期： 双月刊
电话： 024-24686097
Email： sgxb6097@sylu.edu.cn
国内统一刊号(CN)： CN 21-1594/T
国际标准刊号(ISSN)：ISSN 1003-1251