Check failed: ExistsDir(*image_path)

[hm123450]

Hi: I have a video file that is about ten seconds long. After splitting it using ffmpeg, I'm trying to use convert.py, but I keep getting errors related to the file path. I would like to ask if someone could help me with this issue. Thank you very much.

[hm123450]

嗨：我有一个大约十秒钟长的视频文件。使用 ffmpeg 拆分它后，我正在尝试使用 convert.py，但我不断收到与文件路径相关的错误。我想问一下是否有人可以帮助我解决这个问题。谢谢。

python convert.py -s data/
E0222 12:08:50.144686 1326795 logging.cc:56] [option_manager.cc:813] Check failed: ExistsDir(*image_path)
E0222 12:08:50.144773 1326795 option_manager.cc:879] Invalid options provided.
ERROR:root:Feature extraction failed with code 256. Exiting.

Here is my fail information.

[hm123450]

嗨：我有一个大约十秒钟长的视频文件。使用 ffmpeg 拆分它后，我正在尝试使用 convert.py，但我不断收到与文件路径相关的错误。我想问一下是否有人可以帮助我解决这个问题。谢谢。

python convert.py -s data/
E0222 12:08:50.144686 1326795 logging.cc:56] [option_manager.cc:813] Check failed: ExistsDir(*image_path)
E0222 12:08:50.144773 1326795 option_manager.cc:879] Invalid options provided.
ERROR:root:Feature extraction failed with code 256. Exiting.

Here is my fail information.

My file structure :

data -----input.mp4 -----input -----------input_0001.ipg -----------input_0002.jpg -----------etc

[Pumbaa-peng]

try to modify the command to：python convert.py -s data/input

[vivi-gomez]

EDIT** and answer myself. I did move the db file and the original sparce folder inside the distorted folder created by the script and It works now.

---

I have a similar problem. Any light about will be appreciated. My database was generated on OpenCV format (3840x2160) from a gopro camera. The distorted folder is generated by the script but it is empty

-MainFolder ---distorted ----sparce -----(Empty)

---image ----frame_00001.jpg (and so on sequence)

---images_2

....images_4

---images_8

---sparse ----database.db ----0 -----cameras.bin -----images.bin -----points3D.bin -----project.ini

$ python convert.py -s (path to MainFolder) [option_manager.cc:811] Check failed: ExistsDir(*image_path) ERROR: Invalid options provided. ERROR:root:Feature extraction failed with code 256. Exiting.

[y-1111-y]

嗨：我有一个大约十秒钟长的视频文件。使用 ffmpeg 拆分它后，我正在尝试使用 convert.py，但我不断收到与文件路径相关的错误。我想问一下是否有人可以帮助我解决这个问题。谢谢。

python convert.py -s data/
E0222 12:08:50.144686 1326795 logging.cc:56] [option_manager.cc:813] Check failed: ExistsDir(*image_path)
E0222 12:08:50.144773 1326795 option_manager.cc:879] Invalid options provided.
ERROR:root:Feature extraction failed with code 256. Exiting.

Here is my fail information.

My file structure :

data -----input.mp4 -----input -----------input_0001.ipg -----------input_0002.jpg -----------etc

I had the same error with you firstly. And then I read this part. I think you can delete the input.mp4 file.

[yoponchik]

The error is most likely because it cannot find a directory called input

[immutableJackz]

You can fill in the upper-level folder name of the input folder. For example, if your picture is in data/input, you only need to use python convert.py -s data

[kkkwjr]

这里文件夹的名字只能设置为input，我也是无意间看到的，改成input之后就可以运行了

[kkkwjr]

The name of the folder here can only be set to input, I also overlooked it, and after changing it to input, it works!

[dek8v5]

I solved it by using this very specific structure (you have to rename your own dataset to data, then the subfolder to input):

data |___input |_______image1.png |_______image2.png |_______etc

then running the convert.py with: python convert.py -s path/to/data

then the script will create some new directories under data, and it will look like this: data |___input |_______image1.png |_______image2.png |_______etc |___images |_______image1.png |_______image2.png |_______etc |___sparse |______0 |__________camera.bin |__________images.bin |__________points3D.bin |___stereo |______consistency_graphs (empty) |______depth_maps (empty) |______normal_maps (empty) |______fusion.cfg |______path_match.cfg |___distorted |_______sparse |__________0 |_____________camera.bin |_____________images.bin |_____________points3D.bin |_____________project.ini |_database_db | run-colmap-geometric_sh | run-colmap-photometric_sh

[kae1111]

I solved it by using this very specific structure (you have to rename your own dataset to data, then the subfolder to input):

data |___input |_______image1.png |_______image2.png |_______etc

then running the convert.py with: python convert.py -s path/to/data

then the script will create some new directories under data, and it will look like this: data |___input |_______image1.png |_______image2.png |_______etc |___images |_______image1.png |_______image2.png |_______etc |___sparse |______0 |__________camera.bin |__________images.bin |__________points3D.bin |___stereo |______consistency_graphs (empty) |______depth_maps (empty) |______normal_maps (empty) |______fusion.cfg |____path_match.cfg |distorted |_______sparse |__________0 |_____________camera.bin |_____________images.bin |_____________points3D.bin |_____________project.ini |**database_db | run-colmap-geometric_sh |** run-colmap-photometric_sh

it works!thanks!