MIAS-LCEC Toolbox User Guide

Project Navigation

Project Website：MIAS Group

Article：
Online, Target-Free LiDAR-Camera Extrinsic Calibration via Cross-Modal Mask Matching

Toolbox Code：ZWhuang666/MIAS-LCEC (github.com)

Toolbox Demo：bilibili video

Dataset Download：[Google Drive ] [BaiDuYun]

Program Details

1. Start MIAS-LCEC Toolbox

The program consists of two independent pacakges:

No.	Package Name	Function
1	zvision	A C++ package , which is the main UI of the project
2	c3m	A Python package, the main function is to receive command and data from zvision package and complete the segmentation function and cross-modal mask matching function, and then send the matched points pair back to zvision package for further PNP solution.

The zvsion package and c3m package communicate with each other through ros2 topic publishing and subscribing mechanism,there will be some problem if you run both zvision package and c3m package in two computers in a local network since there will be two pubulishing node and subscribe node with the same name in the network. If you need to run zvision and c3m on two or more computers at the same time, please disconnect the computers with the network.

To start MIAS-LCEC Toolbox, you only need to run:

cd ~/MIAS-LCEC/
sh mias_lcec.sh

2. UI Introduction

2.1 control panel

This panel includes checkboxs, inputbox and buttons for general cloud and picture control.

Controls	Function
render Mode checkboxes and inputboxes	choose the cloud rendering mode and set the rendering params.
load CloudPoints Button	load a cloud file and display. then use left/right/mid/mid press/ of the mouse to rotate and zoom the cloud
Rosbag2 Read Button	read the pcd and image in a rosbag2 file, if the checkbox Folder is selected, it can handle all the bag files in a folder, if the checkbox color is selected, all image files will be in RGB mode.
take picture Button	take a cloud picture by the cloud camera and show it in the cloud image window.
load image Button	load an image from a file and show it in the cloud image window
ReadJson Button	read a json file including pics rendering condition and intrinsic/extrinsic parameters, and then set the “cloud camera window” and set the intrinsic/extrinsic parameters of the cloud camera.
SaveJson Button	save the intrinsic/extrinsic parameters of the cloud camera, and the pic rendering conditions to a json file.
Remove All Button	remove all loaded pictures and clouds.