https://github.com/bing-su/my-doctr-ocr
https://github.com/bing-su/my-doctr-ocr
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/bing-su/my-doctr-ocr
- Owner: Bing-su
- License: mit
- Created: 2023-01-10T00:56:00.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-02-27T06:58:59.000Z (over 2 years ago)
- Last Synced: 2025-01-02T15:50:45.066Z (5 months ago)
- Language: Python
- Size: 1.08 MB
- Stars: 2
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# my-doctr-ocr
## Inference
```python
# 불러오기
from doctr.io import DocumentFile
from doctr.models import ocr_predictor, from_hub
# opencv를 쓰기 때문에 영어 이외의 문자가 경로에 들어가면 안됨
img = DocumentFile.from_images(["image.jpg"])
det = from_hub("smartmind/doctr-db_resnet50")
rec = from_hub("smartmind/doctr-vitstr_small-recognition")
# 또는
# rec = from_hub("smartmind/doctr-vitstr_base-recognition")
model = ocr_predictor(det_arch=det, reco_arch=rec)
```
```python
# 추론
result = model(img)
result.show(img)
```
```python
# json으로 결과 보기
>>> result
Output exceeds the size limit. Open the full output data in a text editor
Document(
(pages): [Page(
dimensions=(905, 640)
(blocks): [
Block(
(lines): [
Line(
(words): [
Word(value='상의적', confidence=0.22),
Word(value='인주시면을', confidence=0.66),
Word(value='이르는', confidence=0.27),
Word(value='배ᄉᄉᆫ이ᅵ', confidence=0.16),
]
),
Line(
(words): [Word(value='서울특별시교육청', confidence=0.4)]
),
]
(artefacts): []
),
Block(
(lines): [
Line(
(words): [Word(value='88', confidence=0.21)]
),
...
(artefacts): []
),
]
)]
```
```python
# 모든 한국어는 nfd로 변환되어 있으므로(자모 분리 상태) 필요하면 nfc로 변환해야 함
from unicodedata import normalize
for page in result:
for block in page:
for line in block:
for word in line:
print(normalize("NFC", word.value))
```