FanglinBao / HADAR

This is an LWIR stereo-hyperspectral database to develop HADAR algorithms for thermal navigation. Based on this database, one can develop algorithms for TeX decomposition to generate TeX vision. One can also develop algorithms about object detection, semantic or scene segmentation, optical or scene flow, stereo depth etc. based on TeX vision instead of traditional RGB or thermal vision.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

eMap mapping in GroundTruth

Daepilin opened this issue · comments

Hello Dr. Bao,

Thank you very much for sharing your very interesting work.

While trying to make the example matlab Code work I noticed that, in the ground truth, materials are mapped from some arbitrary (?) shuffled Indices to the order of materials as shown in the material namens DB.

Is there a specific reason why you did this? Is this baked into the scenes (by actually using eg iron emissivity for sky in some scenes) or just done for ground truth generation?

I ask, because I can not get the matlab Code to output expected materials (I teste for example Asphalt and sky pixels in scene 1) , and am wondering if I need to apply some mapping to the outputs?

Hello Daepilin,

Thanks for your interest in our work.
When the scenes in the database were synthesized, each scene had its own subset of the material library, and each scene was synthesized independently by different artists. They were assigned global material indices in post processing for ground truth generation. That leads to the seemingly random shuffled indices.

Thank you very much for the quick response!