# **labelme2yolo**
## 소개
labelme를 이용하여 어노테이션 작업을 한 josn파일을 yolo학습이 가능하도록 txt파일로 변환해주는 기능
## 사용법
Commad line
```
python labelme2yolo.py --json_dir=<s3데이터를 사용한다면 None, 아니면 json파일들이 있는 폴더 경로> --s3_data=<s3데이터를 사용하는지 아닌지에대한 bool값>\
--s3_yaml=<s3접속 정보가 담긴 .yaml파일의 경로> --json_db_table=<사용하려는 json 데이터 베이스의 테이블 명>
```
--s3_data에 입력되는 .yaml파일의 양식 및 필수 키 값
```
# labelme2yolo config
db_region: 'kr_local'
image_data_path: images/bb_seg_data/
json_data_path: json_labels/bb_seg_data/
yolo_data_path: yolo_labels/bb_seg_data/
detect_save_path: 'results/'
# s3 config
s3: True # True or False
bucket: 'bucket-name'
endpoint_url: 'https://kr.object.ncloudstorage.com'
region_name: 'kr-standard'
access_key: 'access-key'
access_token: 'secret-token-key'
```
## 사용 예시
.59 서버에서 작동 시
```
cd jupyter_notebooks/image_utils
python labelme2yolo.py --json_dir=None --s3_data=True --s3_yaml=../data/45_abalone_data/yolo_data/ai_solution_dataset_test.yaml --json_db_table='p1_json_info'
```
# **img_dataset_split**
## 소개
어노테이션 작업이 완료 된 json 혹은 txt 파일을 기준으로 Train, Validation, Test를 원하는 비율에 맞게 나누어서 목록을 생성해주는 기능
## 사용법
Commad line
```
python img_dataset_split.py --yaml_file_path=<s3접속 정보가 담긴 .yaml파일의 경로> --ratio=<train,val,test 비율 기입 예) 8,1,1 > \
--db_config_file_path=<db 접속정보가 담긴 config파일 경로> --save_path=<저장하고 싶은 경로 및 이름> --random_seed=55 \
--json_db_table=<json data의 정보가 있는 테이블 명> --image_db_table=<image data의 정보가 있는 테이블 명> --use_s3=<s3데이터를 사용하는지 아닌지에대한 bool값>
```
--yaml_file_path 입력되는 .yaml파일의 양식 및 필수 키 값
```
# labelme2yolo config
db_region: 'kr_local'
# s3 config
s3: True # True or False
bucket: 'bucket-name'
endpoint_url: 'https://kr.object.ncloudstorage.com'
region_name: 'kr-standard'
access_key: 'access-key'
access_token: 'secret-token-key'
```
```
## 사용 예시
.59 서버에서 작동 시
```
cd jupyter_notebooks/image_utils
python img_dataset_split.py --yaml_file_path=../data/45_abalone_data/yolo_data/ai_solution_dataset_test.yaml --ratio=8,1,1 \
--db_config_file_path=../echoss_query/config/config.yaml --save_path=split_dataset_list.json --random_seed=55 \
--json_db_table=p1_json_info --image_db_table=p1_image_info --use_s3=True
```
내부 함수를 직접사용하고자 할 때
```
sd = SplitDataset(<solution config yaml file path>, (8,1,1), <db config yaml file path>, <json data db table>, <image data db table>, True or False)
sd.save_data_split_json(<save file path>, <random seed : int>)
```
Raw data
{
"_id": null,
"home_page": "",
"name": "echoss-image-utils",
"maintainer": "",
"docs_url": null,
"requires_python": ">3.7",
"maintainer_email": "",
"keywords": "echoss,echoss_image_utils,labelme2yolo,labelme,yolo",
"author": "incheolshin",
"author_email": "incheolshin@12cm.co.kr",
"download_url": "https://files.pythonhosted.org/packages/4b/03/843ba6a75225a19b5a762548230c20794d13aedf2fd4b5c03f5263367673/echoss_image_utils-0.0.2.tar.gz",
"platform": null,
"description": "# **labelme2yolo**\n\n## \uc18c\uac1c\n\nlabelme\ub97c \uc774\uc6a9\ud558\uc5ec \uc5b4\ub178\ud14c\uc774\uc158 \uc791\uc5c5\uc744 \ud55c josn\ud30c\uc77c\uc744 yolo\ud559\uc2b5\uc774 \uac00\ub2a5\ud558\ub3c4\ub85d txt\ud30c\uc77c\ub85c \ubcc0\ud658\ud574\uc8fc\ub294 \uae30\ub2a5\n\n## \uc0ac\uc6a9\ubc95\nCommad line\n```\n python labelme2yolo.py --json_dir=<s3\ub370\uc774\ud130\ub97c \uc0ac\uc6a9\ud55c\ub2e4\uba74 None, \uc544\ub2c8\uba74 json\ud30c\uc77c\ub4e4\uc774 \uc788\ub294 \ud3f4\ub354 \uacbd\ub85c> --s3_data=<s3\ub370\uc774\ud130\ub97c \uc0ac\uc6a9\ud558\ub294\uc9c0 \uc544\ub2cc\uc9c0\uc5d0\ub300\ud55c bool\uac12>\\\n --s3_yaml=<s3\uc811\uc18d \uc815\ubcf4\uac00 \ub2f4\uae34 .yaml\ud30c\uc77c\uc758 \uacbd\ub85c> --json_db_table=<\uc0ac\uc6a9\ud558\ub824\ub294 json \ub370\uc774\ud130 \ubca0\uc774\uc2a4\uc758 \ud14c\uc774\ube14 \uba85>\n```\n--s3_data\uc5d0 \uc785\ub825\ub418\ub294 .yaml\ud30c\uc77c\uc758 \uc591\uc2dd \ubc0f \ud544\uc218 \ud0a4 \uac12\n```\n # labelme2yolo config\n db_region: 'kr_local'\n image_data_path: images/bb_seg_data/\n json_data_path: json_labels/bb_seg_data/\n yolo_data_path: yolo_labels/bb_seg_data/\n detect_save_path: 'results/'\n\n # s3 config\n s3: True # True or False\n\n bucket: 'bucket-name'\n endpoint_url: 'https://kr.object.ncloudstorage.com'\n region_name: 'kr-standard'\n access_key: 'access-key'\n access_token: 'secret-token-key'\n```\n\n## \uc0ac\uc6a9 \uc608\uc2dc\n.59 \uc11c\ubc84\uc5d0\uc11c \uc791\ub3d9 \uc2dc\n```\n cd jupyter_notebooks/image_utils\n \n python labelme2yolo.py --json_dir=None --s3_data=True --s3_yaml=../data/45_abalone_data/yolo_data/ai_solution_dataset_test.yaml --json_db_table='p1_json_info'\n\n```\n\n\n# **img_dataset_split**\n\n## \uc18c\uac1c\n\uc5b4\ub178\ud14c\uc774\uc158 \uc791\uc5c5\uc774 \uc644\ub8cc \ub41c json \ud639\uc740 txt \ud30c\uc77c\uc744 \uae30\uc900\uc73c\ub85c Train, Validation, Test\ub97c \uc6d0\ud558\ub294 \ube44\uc728\uc5d0 \ub9de\uac8c \ub098\ub204\uc5b4\uc11c \ubaa9\ub85d\uc744 \uc0dd\uc131\ud574\uc8fc\ub294 \uae30\ub2a5\n\n## \uc0ac\uc6a9\ubc95\nCommad line\n```\n python img_dataset_split.py --yaml_file_path=<s3\uc811\uc18d \uc815\ubcf4\uac00 \ub2f4\uae34 .yaml\ud30c\uc77c\uc758 \uacbd\ub85c> --ratio=<train,val,test \ube44\uc728 \uae30\uc785 \uc608) 8,1,1 > \\\n --db_config_file_path=<db \uc811\uc18d\uc815\ubcf4\uac00 \ub2f4\uae34 config\ud30c\uc77c \uacbd\ub85c> --save_path=<\uc800\uc7a5\ud558\uace0 \uc2f6\uc740 \uacbd\ub85c \ubc0f \uc774\ub984> --random_seed=55 \\\n --json_db_table=<json data\uc758 \uc815\ubcf4\uac00 \uc788\ub294 \ud14c\uc774\ube14 \uba85> --image_db_table=<image data\uc758 \uc815\ubcf4\uac00 \uc788\ub294 \ud14c\uc774\ube14 \uba85> --use_s3=<s3\ub370\uc774\ud130\ub97c \uc0ac\uc6a9\ud558\ub294\uc9c0 \uc544\ub2cc\uc9c0\uc5d0\ub300\ud55c bool\uac12>\n```\n\n--yaml_file_path \uc785\ub825\ub418\ub294 .yaml\ud30c\uc77c\uc758 \uc591\uc2dd \ubc0f \ud544\uc218 \ud0a4 \uac12\n```\n # labelme2yolo config\n db_region: 'kr_local'\n\n # s3 config\n s3: True # True or False\n\n bucket: 'bucket-name'\n endpoint_url: 'https://kr.object.ncloudstorage.com'\n region_name: 'kr-standard'\n access_key: 'access-key'\n access_token: 'secret-token-key'\n```\n```\n\n## \uc0ac\uc6a9 \uc608\uc2dc\n.59 \uc11c\ubc84\uc5d0\uc11c \uc791\ub3d9 \uc2dc\n```\n cd jupyter_notebooks/image_utils\n \n python img_dataset_split.py --yaml_file_path=../data/45_abalone_data/yolo_data/ai_solution_dataset_test.yaml --ratio=8,1,1 \\\n --db_config_file_path=../echoss_query/config/config.yaml --save_path=split_dataset_list.json --random_seed=55 \\\n --json_db_table=p1_json_info --image_db_table=p1_image_info --use_s3=True\n```\n\n\ub0b4\ubd80 \ud568\uc218\ub97c \uc9c1\uc811\uc0ac\uc6a9\ud558\uace0\uc790 \ud560 \ub54c\n```\n sd = SplitDataset(<solution config yaml file path>, (8,1,1), <db config yaml file path>, <json data db table>, <image data db table>, True or False)\n \n sd.save_data_split_json(<save file path>, <random seed : int>)\n```\n",
"bugtrack_url": null,
"license": "",
"summary": "echoss AI Bigdata Solution - image utils[image dataset split, labelme2yolo format]",
"version": "0.0.2",
"project_urls": null,
"split_keywords": [
"echoss",
"echoss_image_utils",
"labelme2yolo",
"labelme",
"yolo"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "58b46957c5d3ac3936786b32c2b27fb923363c87ee1ff4de09d82d9b561d9879",
"md5": "fca6e80848f64cd82a50cebd09202f35",
"sha256": "1fd6c57cc505378a822d9f0fd63d49cc1cbf2b9bd6c7afb622f408b91a8175c3"
},
"downloads": -1,
"filename": "echoss_image_utils-0.0.2-py3-none-any.whl",
"has_sig": false,
"md5_digest": "fca6e80848f64cd82a50cebd09202f35",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">3.7",
"size": 14731,
"upload_time": "2024-01-08T06:21:59",
"upload_time_iso_8601": "2024-01-08T06:21:59.460807Z",
"url": "https://files.pythonhosted.org/packages/58/b4/6957c5d3ac3936786b32c2b27fb923363c87ee1ff4de09d82d9b561d9879/echoss_image_utils-0.0.2-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "4b03843ba6a75225a19b5a762548230c20794d13aedf2fd4b5c03f5263367673",
"md5": "e950d2a0b0f6835ad8b9fdb370e58706",
"sha256": "a9a2beb384ccb24aab8c1e68d4a1c53d95547ea48406619cb3bf03651e89cfcf"
},
"downloads": -1,
"filename": "echoss_image_utils-0.0.2.tar.gz",
"has_sig": false,
"md5_digest": "e950d2a0b0f6835ad8b9fdb370e58706",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">3.7",
"size": 12543,
"upload_time": "2024-01-08T06:22:01",
"upload_time_iso_8601": "2024-01-08T06:22:01.004493Z",
"url": "https://files.pythonhosted.org/packages/4b/03/843ba6a75225a19b5a762548230c20794d13aedf2fd4b5c03f5263367673/echoss_image_utils-0.0.2.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-01-08 06:22:01",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "echoss-image-utils"
}
JSON