Calibration error - calib_intri.py v0.2

[IvanHalimau]

python apps/calibration/calib_intri.py 0_input/project_calib --step 5 -this command line

result:
(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/calib_intri.py 0_input/project_calib --step 5
usage: calib_intri.py [-h] [--image IMAGE] [--ext {.jpg,.png}] [--num NUM] [--sample SAMPLE] [--share_intri]
[--remove]
path
calib_intri.py: error: unrecognized arguments: --step 5

also delete "--step 5" - doesn't write an error, but writes this:
result:
(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/calib_intri.py 0_input/project_calib
read: 100%|████████████████████████████████████████████████████████████████████████| 570/570 [00:00<00:00, 4671.08it/s]

Camera 1: 570 frames

and so it costs a long time for 10 minutes. The github says python3 instead of python, but it doesn't help, maybe I'm doing something wrong?

[chingswy]

Hello, you can add flag --sample 200 to calibrate with 200 images.

[IvanHalimau]

Hello, you can add flag --sample 200 to calibrate with 200 images.

thanks!

[IvanHalimau]

(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/calib_intri.py 0_input/project_calib --sample 200
read: 100%|█████████████████████████████████████████████████████████████████████████| 570/570 [00:01<00:00, 288.16it/s]

Camera 1: 570 frames

Hey! Not working :(
Maybe not work GPU? I'm install Python 3.7+ and Nvidia CUDA 11. Before updating to v0.2 its working. I'm fresh install now easymocap.

LOG:
(easymocap) C:\MOCAP\EasyMocap>conda activate easymocap

(easymocap) C:\MOCAP\EasyMocap>python scripts/preprocess/extract_video.py 0_input/project_calib --no2d
1.mp4 : 100%|████████████████████████████████████████████████████████████████████| 570/570 [00:26<00:00, 21.28it/s]
2.mp4 : 100%|████████████████████████████████████████████████████████████████████| 549/549 [00:25<00:00, 21.31it/s]
cameras: 1 2

(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/detect_chessboard.py 0_input/project_calib --out output/calibration --pattern 4,3 --grid 0.135
Create chessboard (4, 3)
create template chessboard: 100%|████████████████████████████████████████████████| 1119/1119 [00:00<00:00, 1180.60it/s]
100%|██████████████████████████████████████████████████████████████████████████████| 1119/1119 [03:04<00:00, 6.07it/s]

(easymocap) C:\MOCAP\EasyMocap>
(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/calib_intri.py 0_input/project_calib --step 5
usage: calib_intri.py [-h] [--image IMAGE] [--ext {.jpg,.png}] [--num NUM] [--sample SAMPLE] [--share_intri]
[--remove]
path
calib_intri.py: error: unrecognized arguments: --step 5

(easymocap) C:\MOCAP\EasyMocap>
(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/calib_intri.py 0_input/project_calib --step 5
usage: calib_intri.py [-h] [--image IMAGE] [--ext {.jpg,.png}] [--num NUM] [--sample SAMPLE] [--share_intri]
[--remove]
path
calib_intri.py: error: unrecognized arguments: --step 5

(easymocap) C:\MOCAP\EasyMocap>python apps/calibration/calib_intri.py 0_input/project_calib --sample 200
read: 100%|█████████████████████████████████████████████████████████████████████████| 570/570 [00:01<00:00, 288.16it/s]

Camera 1: 570 frames

[IvanHalimau]

I'm reinstalling Nvidia CUDA 11.7 - no changes :(

[theunit187]

I have the same issue.
python apps/calibration/calib_intri.py 0_input/project_calib --step 5
It can't recognize --step 5, and if you remove it, it prints:
>> Camera 1: 428 frames
...and is getting stuck. I've left it open for hours, nothing happens. I've tried 200 samples, but the result is the same.

[chingswy]

Can you try to calibrate with --sample 10 or --sample 20?

[Lilien86]

I have exactly the same issue; the arguments --sample or --step don't work. In every case, the file 'calib_intri.py' runs with all frames of my video, which makes the process take a very long time to finish

[Lilien86]

to choice the number of images that we want, we can put the argument "--num" wich determinate the number of image treaty.
----python apps/calibration/calib_intri.py .\0_input\project_calib\ --num 100
logically when we put the argument "--sample" after "--num" this should ensure that of these X immages, only X are chosen and in a random manner.
----python apps/calibration/calib_intri.py .\0_input\project_calib\ --num 100 --sample 50