Hey Community,
I would like to generate a hdr image from multiple ldr images. My goal is to map pixel values (0-255) to W/m² solarpower via machine learning. I use a raspberry pi with raspberry hq cam (imx477). Unfortunalty I cant find any information about * genHDR (from http://www.anyhere.com/)… I would like to understand the mathematics behind the transformation. Could someone please provide further information or give me any tips for my project.
Happy Greetings
Paul
Hi Paul,
You are already aware of hdrgen (Linux) and photosphere (Mac). Here is the download and documentation.
If you’d like a non-Radiance solution, here is a comprehensive explanation of the process, although I don’t know if it’s quite the method that @Greg_Ward used.
The basic method used in Photosphere and hdrgen is by Mitsunaga and Nayar:
T. Mitsunaga and S. K. Nayar, Radiometric Self Calibration, Proc. IEEE Conference on Computer Vision and Pattern Recognition, June 1999.
I have added a number of enhancements, most of which are described in chapter 5 of the second edition of Reinhard et al.
Cheers,
-Greg
