CRAFT pytorch
1.0.0
공예 텍스트 탐지기의 공식 Pytorch 구현 | 종이 | 사전 예방 모델 | 보충
Youngmin Baek, Bado Lee, Dongyoon Han, Sangdoo Yun, Hwalsuk Lee.
Clova AI Research, Naver Corp.
각 캐릭터 영역과 문자 간의 친화력을 탐색하여 텍스트 영역을 효과적으로 감지하는 크래프트 텍스트 검출기의 Pytorch 구현. 텍스트의 경계 상자는 임계 값 문자 영역 및 친 화성 점수 후 이진 맵에서 최소 경계 사각형을 찾아서 얻습니다.

2019 년 6 월 13 일 : 2019 년 6 월 13 일 : 2019 년 7 월 20 일 : 다각형 결과에 대한 후 처리 추가 추가 2019 년 9 월 28 일 : IC15 및 Link Refiner의 훈련 된 모델이 추가되었습니다.
pip install -r requirements.txt
교육 코드는이 저장소에 포함되지 않으며 IP 이유에 대한 전체 교육 코드를 공개 할 수 없습니다.
| 모델 이름 | 중고 데이터 세트 | 언어 | 목적 | 모델 링크 |
|---|---|---|---|---|
| 일반적인 | SynthText, IC13, IC17 | ENG + MLT | 일반적인 목적을 위해 | 딸깍 하는 소리 |
| IC15 | SynthText, IC15 | 잉그 | IC15 만 | 딸깍 하는 소리 |
| LinkRefiner | CTW1500 | - | 일반 모델과 함께 사용됩니다 | 딸깍 하는 소리 |
python test.py --trained_model=[weightfile] --test_folder=[folder path to test images]
결과 이미지 및 소셜 맵은 기본적으로 ./result 로 저장됩니다.
--trained_model : 사전 예방 모델--text_threshold : 텍스트 신뢰 임계 값--low_text : 텍스트 로우 바운드 점수--link_threshold : 링크 신뢰 임계 값--cuda : 추론에 Cuda를 사용하십시오 (기본값 : true)--canvas_size : 추론을위한 최대 이미지 크기--mag_ratio : 이미지 확대 비율--poly : 다각형 유형 결과를 활성화합니다--show_time : 처리 시간을 보여줍니다--test_folder : 입력 이미지를위한 폴더 경로--refine : Sentense 수준 데이터 세트에 Link Refiner를 사용하십시오--refiner_model : 사전 상환 정유소 모델 @inproceedings{baek2019character,
title={Character Region Awareness for Text Detection},
author={Baek, Youngmin and Lee, Bado and Han, Dongyoon and Yun, Sangdoo and Lee, Hwalsuk},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
pages={9365--9374},
year={2019}
}
Copyright (c) 2019-present NAVER Corp.
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.