论文 - Githubissues

maomao0220 commented 4 weeks ago

请问，可以提供论文吗

Yixin-F commented 4 weeks ago

https://ieeexplore.ieee.org/abstract/document/10419066

maomao0220 commented 4 weeks ago

大佬你好，在运行better_fastlio的建图功能时，并开启了动态去除功能，但是运行结果显示动态点云去除不完全，很多点保留在了地图中，请问需要调节哪些参数？其中源代码中的雷达高度已经改成实际使用的高度微信图片_20240815180427 参数如下： `common: lid_topic: "/front_main_lidar_driver/valid_cloud" imu_topic: "/ouster1_front/imu" camera_topic: "/usb_cam/image_raw" time_sync_en: false time_offset_lidar_to_imu: 0.0

rootDir: ""
savePCD: false
savePCDDirectory: "PCDs/"        
saveSCD: false 
saveSCDDirectory: "SCDs/"
saveLOG: false
saveLOGDirectory: "LOG/"
map_save_en: true  # global map
pcd_save_interval: 1                 # how many LiDAR frames saved in each pcd

preprocess: lidar_type: 2 # 1 Velodyne LiDAR 2: Ouster 3: RS scan_line: 128 blind: 1.5 feature_enabled: 0 # bool point_filter_num: 6 # bool scan_rate: 10 time_unit: 3

camera: camera_en: false camera_external: [ 0.000000, -1.000000, 0.000000, 0.000000, 0.000000, 0.000000, -1.000000, 0.300000, 1.000000, 0.000000, 0.000000, 0.000000, 0.000000, 0.000000, 0.000000, 1.000000] #相机2雷达外参 camera_internal: [516.235840, 0.000000, 641.125888, 0.000000, 0.000000, 533.185608, 338.725370, 0.000000, 0.000000, 0.000000, 1.000000, 0.000000] #相机内参

mapping: acc_cov: 0.1 gyr_cov: 0.1 b_acc_cov: 0.0001 b_gyr_cov: 0.0001 fov_degree: 180.0 det_range: 100.0 cube_len: 1000
extrinsic_est_en: false # true: enable the online estimation of IMU-LiDAR extrinsic mappingSurfLeafSize: 1.0 # 0.2 keyframeAddingDistThreshold: 1.0 keyframeAddingAngleThreshold: 0.2 extrinsic_T: [ -0.006253, 0.011775, -0.007645] # lidar-imu extrinsic_R: [ 1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0]

ikdtree: max_iteration: 3 recontructKdTree: false # if path is too large, the rviz will crash, but default is false kd_step: 30 filter_size_map_min: 1.0 # 0.5

loop: loopClosureEnableFlag: false loopClosureFrequency: 1.0 historyKeyframeSearchRadius: 10.0 historyKeyframeSearchTimeDiff: 30.0 historyKeyframeSearchNum: 2 historyKeyframeFitnessScore: 0.2

segment: ground_en: false sensor_height: 2.5 tollerance_en: false # error z_tollerance: 2.0 rotation_tollerance: 0.2

publish: path_en: true scan_publish_en: true # false: close all the point cloud output dense_publish_en: false # false: low down the points number in a global-frame point clouds scan. scan_bodyframe_pub_en: false # true: output the point cloud scans in IMU-body-frame

visualization: globalMapVisualizationSearchRadius: 10.0 globalMapVisualizationPoseDensity: 0.5 globalMapVisualizationLeafSize: 0.2 visulize_IkdtreeMap: false` 请问该如何调节参数？非常期待您的回复，感谢

Yixin-F commented 4 weeks ago

你好，感谢你的使用。有些dynamic removal 的 parameter我写在better_fastlio2/include/dynamic-remove/tgrs.h里了，没有放在yaml里，你可以修改以下#define HD_RATIO (float)0.7 ，调为更高。因为better fastlio2中的dynamic removal跟object-level update有许多函数是共用的，可能会有些功能上的冲突，如果你修改HD_RATIO之后还不行，我需要进一步debug以下，谢谢。

maomao0220 commented 3 weeks ago

试了一下您说的建议，增加了HD_RATIO的数值，最终静态建图结果还是会保留部分动态点云，方便的话，可以发给您相关bag包，使用的是os1-128雷达以及内置imu数据

maomao0220 commented 3 weeks ago

你好，根据您提供的论文链接，我下载不了，可否提供下载后的论文？

Yixin-F / DR-Using-SCV-OD

论文 #1