How to ensure correct exposures of ldr image

hdri
1

Dear all,
I read Pierson’s tutorial and I am very confused about choosing the right exposure. I put the 13 bracketed photos in Google Drive.
https://drive.google.com/drive/folders/1O2aoS9kK9_82e9FX43yyALCOouNz88XG?usp=sharing

I wrote the following Python program to automatically identify valid exposures, but after traversing thousands of images, I did not find even one image with valid exposure (that is, R, G, B values ​​between 27 and 228), so that when I called evalglare to analyze the synthesized HDR image, the result was ridiculous. I want to know if there is any remedy?

dddca5dbe9f1ad36399e9955ed37def
dddca5dbe9f1ad36399e9955ed37def1116×740 65.4 KB

3271587a96ec71777dd3b61e466b72e
3271587a96ec71777dd3b61e466b72e930×648 57.7 KB

I have communicated with Greg a lot and he helped me a lot, but he didn’t have a specific solution to my problem, so I would like to ask everyone in the forum to help me, thank you.

2

Hi @HaohangLi,

Welcome to the Radiance forum. If your goal is to determine whether the image contains some useful pixels in the range 27 < pixel value < 228, then you should be testing np.all, not np.any.