SLAM Hive: A SLAM Evaluation Benchmark System ()

Wiki Link
Mapping Task Configuration 2470 Details Information

mapping task ID: 3490 |
  
mapping task ID: 3491 |
  


Algorihtm & Dataset

Algorithm Selection

Algorithm ImageTag: slambench3-orb-slam2-stereo
Algorihtm name: slambench3

Dataset Selection:

Dataset Sequence Name: slambench3_MH_01_easy
Dataset name: slambench3


Configuration Information

Mapping Task Configuration Name: slambench3-orb-slam2-stereo+slambench3_MH_01_easy
Mapping Task Configuration Description:

slambench3 runs orbslam2 stereo on MH_01_easy.
We provide some parameters(provided by slambench) to set, but we are not really sure the true meaning of each parameter.


Corresponding Parameters Infomation

ID Name Description parameter Type Algo/Dataset Name Key Value Type Value
345 ORB+nFeatures
Number of features per image
Algorithm slambench3 --max-features int 1000
346 ORB+scaleFactor
Scale factor between levels in the scale pyramid
Algorithm slambench3 --scale-factor float 1.2
347 ORB+nLevels
Number of levels in the scale pyramid
Algorithm slambench3 --scale-levels int 8
348 ORB+iniThFAST
Firstly we impose iniThFAST. If no corners are detected we impose a lower value minThFAST
Algorithm slambench3 --initial-fast-threshold int 20
349 ORB+minThFAST
You can lower these values if your images have low contrast
Algorithm slambench3 --second-fast-threshold int 7
350 general+image_frequency
frequency of image data
Dataset slambench3 -fps int 20
449 general+ThDepth
extrinsic
Dataset euroc --depth-threshold int 40
624 slambench3+realtime-mode
select slambench running mode:
  - 0: not real time
  - 1: real time
General parameter slambench3 --realtime-mode int 0