hi, we use custom data. But make failed by NEUS's guidance.

[As-Time-Goes-By]

[ 50%] Building CXX object CMakeFiles/calibration.dir/calibration.cpp.o [100%] Linking CXX executable calibration /usr/bin/ld: cannot find -lcnpy collect2: error: ld returned 1 exit status CMakeFiles/calibration.dir/build.make:137: recipe for target 'calibration' failed make[2]: *** [calibration] Error 1 CMakeFiles/Makefile2:82: recipe for target 'CMakeFiles/calibration.dir/all' failed make[1]: *** [CMakeFiles/calibration.dir/all] Error 2 Makefile:90: recipe for target 'all' failed make: *** [all] Error 2

[yuedajiong]

say thanks to me, and follow me. (if you still can not reconstrcut your object, please feel free to mail me: [email protected])

0. before step by step, some tips about how to solve problem without friendly help, like me. a) find similar projects, similar camera pose computing. in nerf, almost all of them used colmap, excepts a few camera-pose-free algorithms, such as: nope, four, nerf--, etc. b) from the classical implementation, you can find some branchs, that means there are many many researcher forked and modified, including this type of data-preprocessing issue.

1. step by step a) install colmap, you can download and install windows version, I will copy my steps, at the end of this answer. b) download that 'DOG' data from https://github.com/xrr-233/Textured-NeuS. ([think back my suggestion #0, almost all of problems have been resolved by others.) c) copy code, under preprocess_custom_data, from https://github.com/xrr-233/Textured-NeuS. d) run two steps: *.py. e) bingo. :-)

0.txt

[As-Time-Goes-By]

@yuedajiong. Thanks for your help. I will try by your steps.