search

rubengooj / pl-svo

This code contains an algorithm to compute monocular visual odometry by using both point and line segment features, based on the open source version of SVO.
GNU General Public License v3.0
241 stars 73 forks source link

could you share your config file on EuRoC MAV dataset #1

Open extr15 opened 7 years ago

extr15 commented 7 years ago

Hi,your PL-SVO looks nice and I want to test it on EuRoC_MAV/V1_01_easy, but the tracking just lost in the several beginning images. I have modified the dataset_params.yaml using the camera intrinsic parameters of EuRoC_MAV/V1_01_easy. Do I miss anything or could you share your config file when testing on some datasets? such as a file called dataset_params_EuRoC_V1_01_easy.yaml or dataset_params_ICL_NUIM_seq_02.yaml. Btw, in sceneRepresentation.cpp, some files like aux/legend.png, aux/legend_comp.png are loaded but no such files in this repo, could you add them? Thanks!

cumtchenchang commented 6 years ago

@can you tell me how run this sucessfully? I put dataset_params.yaml into the mav0, than correct the last line to images_subfolder: cam0/data/ than run ./bin/run_pipeline /home/cc/data/mav0 I tried many ways, but error segmentation fault (core dumpled) still on

2016cxg commented 5 years ago

@can you tell me how run this sucessfully? I put dataset_params.yaml into the mav0, than correct the last line to images_subfolder: cam0/data/ than run ./bin/run_pipeline /home/cc/data/mav0 I tried many ways, but error segmentation fault (core dumpled) still on

As is told in the statement where refers to the sequence folder relative to the environment variable ${DATASETS_DIR} that must be previously set from Readme.md, not an absolute path which goes like /home/cc/data/mav0 is needed, that is simply the name of your folder contain your image files. Maybe my other comments could offer a help

xiaozhi12345678 commented 5 years ago

@2016cxg Did you run sucessfully on Euroc dataset？