Author : Sai Siddartha Maram
Lab : Research Assistant Transportation Lab, GeorgiaTech
Professor and Supervisor : Dr. Yi-Chang (James) Tsai
The insight tool provides an opportunity to analyse and study the behaviour of retro-intensity captured form LiDAR on traffic signs. Prominent features of the tool as for the latest push include
- Statistical analysis and sign replacement strategies
- 3D visualiation of the LiDAR signs based on retro
- Identifying areas of damage on the sign
- Relation between color and its effect on retro intensity
- Study specific areas of the sign using Lasso selection technique
- Behaviour of sign and its relation with retro intensity with age and persistent energy absorbtion
From various statistic distributions the tool uses median to give recommendations, this is because of spurious points on the boundary will disturb the mean.
The tool loads the data (user selected sign id) from the lidar inventory and populates histogram and the corresponding 3D heights based on the retro values
The tool allows us to get insight on the spurious points and also generate insight into different points based on retro values
The lasso analysis allows user to draw ovIndia1983er certain points they are interested to study, this allows us to study the behaviour and trends of retrointensity based on age, color and region
Place your data in the Data folder. Data consists of an .csv sheet named signs.csv. The GUI tool uses pX,pY,Retro,Color values for each sign_id. Make sure you fill these coloumns correctly as per the data you have collected. Please note this application is developed based on the output of the colorizer if you plan to use different data, make sure you adjust your data coloumn names accordingly. Do not use the GPS coordinates to plot the graphs, since the points will be highly indisiguisable on 3d plots so make sure you use only pX and pY.
(Change line number 52, if you want to anyother sign csv sheet)
In the code section, you will find the UI_with_lasso.py, this is the core algorithm and has the GUI code within. Install the dependencies required and run the.
python UI_with_lasso.py
Post which all the tools can be used from the GUI.