Spaces:
Runtime error
Runtime error
File size: 2,536 Bytes
e4b7f54 2f465bd e4b7f54 8029506 2f465bd e4b7f54 2f465bd 8029506 e4b7f54 2f465bd e4b7f54 9adb515 0dbdcad 9adb515 2f465bd 0dbdcad 8029506 0dbdcad 2f465bd 8029506 0dbdcad e4b7f54 2f465bd 8029506 4b32c6f 8029506 4b32c6f 8029506 4b32c6f 8029506 2f465bd e4b7f54 8029506 e4b7f54 8029506 2f465bd 8029506 2f465bd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 |
import os
import io
import easyocr
import numpy as np
from fastapi import FastAPI, UploadFile, File
from fastapi.responses import JSONResponse
from PIL import Image
from pdf2image import convert_from_bytes
from concurrent.futures import ThreadPoolExecutor
# =========================
# EasyOCR config
# =========================
MODEL_DIR = "/app/.EasyOCR"
USER_NET_DIR = os.path.join(MODEL_DIR, "user_network")
os.makedirs(MODEL_DIR, exist_ok=True)
os.makedirs(USER_NET_DIR, exist_ok=True)
# ✅ preload reader with cached models
reader = easyocr.Reader(
['en', 'hi'], # langs (reduce if only English needed)
model_storage_directory=MODEL_DIR,
user_network_directory=USER_NET_DIR,
download_enabled=False # 🚫 block downloads at runtime
)
# =========================
# FastAPI app
# =========================
app = FastAPI()
@app.get("/")
async def root():
return {"message": "OCR API is running on Hugging Face 🚀"}
def run_ocr_on_image(image: Image.Image):
"""Convert PIL → numpy and run OCR"""
image_np = np.array(image)
results = reader.readtext(image_np)
text_results = []
for bbox, text, prob in results:
# ✅ convert bbox coords to plain Python floats
bbox_py = [[float(x), float(y)] for x, y in bbox]
text_results.append({
"bbox": bbox_py,
"text": str(text),
"confidence": float(prob)
})
return text_results
@app.post("/ocr")
async def ocr(file: UploadFile = File(...)):
try:
contents = await file.read()
# Detect file type
if file.filename.lower().endswith(".pdf"):
# ✅ Convert PDF to images
pages = convert_from_bytes(contents)
# ✅ Run OCR in parallel
text_results = []
with ThreadPoolExecutor() as executor:
results_list = list(executor.map(run_ocr_on_image, pages))
for i, page_results in enumerate(results_list, start=1):
text_results.append({
"page": i,
"results": page_results
})
return JSONResponse(content={"pdf_results": text_results})
else:
# ✅ Normal image case
image = Image.open(io.BytesIO(contents))
text_results = run_ocr_on_image(image)
return JSONResponse(content={"results": text_results})
except Exception as e:
return JSONResponse(content={"error": str(e)}, status_code=500)
|