kou1okada/libelas

LIBELAS: Library for Efficient Large-scale Stereo Matching (Unofficial fork)

libelas's Issues

how to input png or jpg stereo images for disparity map calculation

When I try to input .png stereo images pair instead of .pgm, it gave me this error:
"/elas img/left_1.png img/right_1.png
Processing: img/left_1.png, img/right_1.png
ERROR: Could not read file img/left_1.png
terminate called after throwing an instance of 'pnm_error'
Aborted"
Can you help me in this regard

Tips for resolving ERROR: Need at least 3 support points?

Hello, I have two images that do not have a whole lot of features. When I try to generate the disparity map, I get an error saying "ERROR: Need at least 3 support points!". Do you have any tips on how to resolve this. Are there any internal parameters I could try to tweak?

I am very new to this so I apologise if there is an obvious answer to this

Thanks!

MSVC build error

With msvc 19.28 we get error C2371 at typedef of each header file. __int8 can be signed or unsigned. Although signed is the default (according to msdn) a compiler option might be making unsigned the default. Explicitly signing all these types fixed this issue for me

improve the result

Thanks for your work but I didn't get desired result from this lib.
I have two pgm images as following


(The upload pictures are png version seems github issue cannot allow me to upload pgm version but in the code I use .pgm picture)
The result is like the follwing


I just use the original demo main.cpp. Anyway to improve it?
Thanks!

scaling of the result

Hi, I tried this lib and it worked fine, but only one problem occurred.
When I converted the output disparity map to depth value expressed in current camera frame, the depth seems relatively short. What I expected is something around 2 meters, but here I got were around 0.3 meters.
I also tested my rectification and converting disparity to depth with OpenCV, so I can say settings from disparity to depth is correct.
My question is, does Libelas do some scaling on the output?