search

yanyan-li / PlanarSLAM

A RGB-D SLAM system for structural scenes, which makes use of point-line-plane features and the Manhattan World assumption.
381 stars 85 forks source link

针对奥比中光深度相机的平面调参技巧，以及living_room数据集格式问题 #2

Closed why-freedom closed 3 years ago

why-freedom commented 3 years ago

请问使用实际相机测试发现平面上的点不如论文中的密集。这种情况应该怎么调节参数？

环境：ubuntu16.04 相机：奥比中光astrapro 深度相机

why-freedom commented 3 years ago

请问living_room数据格式和tum不一样，我尝试自己写associte将其关联，但是名称并不是时间戳，所以会报错。您可以提供相关的工具包或者数据吗

yanyan-li commented 3 years ago

Hi, Thanks for testing our system.

  1. In order to improve the tracking efficiency, a threshold is set to filter the redundant points on the plane, which can be changed to generate dense planes. Frame.cc: pcl::VoxelGrid voxel; voxel.setLeafSize( );
yanyan-li commented 3 years ago

请问living_room数据格式和tum不一样，我尝试自己写associte将其关联，但是名称并不是时间戳，所以会报错。您可以提供相关的工具包或者数据吗 If you have the correct rgb.txt and depth.txt files, the TUM RGB-D [tool] (https://vision.in.tum.de/data/datasets/rgbd-dataset/tools) can help you generate a correct assocition.txt file. Another simple way is to swap the names of depth and rgb forders.