Create app.py

app.py

@@ -0,0 +1,162 @@
+import gradio as gr
+from rapidocr_onnxruntime import RapidOCR
+import cv2
+import numpy as np
+from fastapi import FastAPI, UploadFile, File, Form, HTTPException
+from pydantic import BaseModel
+from typing import Optional
+import io
+from PIL import Image
+import uvicorn
+
+engine = RapidOCR()
+
+# Tạo FastAPI app
+fastapi_app = FastAPI(title="OCR API", description="API for OCR recognition using RapidOCR")
+
+def process_ocr(image: np.ndarray, use_det: bool, use_cls: bool, use_rec: bool) -> list:
+    """Xử lý OCR và trả về kết quả"""
+    img_bgr = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
+    result, _ = engine(img_bgr, use_det=use_det, use_cls=use_cls, use_rec=use_rec)
+    
+    if not result:
+        return []
+    
+    texts = []
+    for item in result:
+        if len(item) == 3:
+            box, text, score = item
+            texts.append({
+                "text": text,
+                "confidence": float(score),
+                "bbox": box.tolist() if hasattr(box, 'tolist') else box
+            })
+        elif len(item) == 2:
+            _, text = item
+            texts.append({
+                "text": str(text),
+                "confidence": None,
+                "bbox": None
+            })
+    
+    return texts
+
+def recognize_text_gradio(image, use_det, use_cls, use_rec):
+    """Hàm cho Gradio interface"""
+    if image is None:
+        return "No image uploaded"
+    
+    results = process_ocr(image, use_det, use_cls, use_rec)
+    
+    if not results:
+        return "No text found"
+    
+    output_lines = []
+    for item in results:
+        if item['confidence']:
+            output_lines.append(f"{item['text']} (score: {item['confidence']:.3f})")
+        else:
+            output_lines.append(item['text'])
+    
+    return "\n".join(output_lines)
+
+# FastAPI Endpoints
+@fastapi_app.post("/ocr")
+async def ocr_endpoint(
+    file: UploadFile = File(..., description="Image file to process"),
+    use_det: bool = Form(True, description="Use detection"),
+    use_cls: bool = Form(True, description="Use classification"),
+    use_rec: bool = Form(True, description="Use recognition")
+):
+    """
+    OCR endpoint that accepts image file upload
+    """
+    # Check file type
+    if not file.content_type.startswith('image/'):
+        raise HTTPException(status_code=400, detail="File must be an image")
+    
+    try:
+        # Read image file
+        contents = await file.read()
+        image = Image.open(io.BytesIO(contents))
+        
+        # Convert to RGB numpy array
+        if image.mode != 'RGB':
+            image = image.convert('RGB')
+        img_np = np.array(image)
+        
+        # Process OCR
+        results = process_ocr(img_np, use_det, use_cls, use_rec)
+        
+        return {
+            "success": True,
+            "texts": [item["text"] for item in results],
+            "details": results,
+            "num_texts": len(results)
+        }
+    
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error processing image: {str(e)}")
+
+@fastapi_app.get("/health")
+async def health_check():
+    """Health check endpoint"""
+    return {"status": "healthy", "service": "OCR API"}
+
+class OCRURLRequest(BaseModel):
+    url: str
+    use_det: Optional[bool] = True
+    use_cls: Optional[bool] = True
+    use_rec: Optional[bool] = True
+
+@fastapi_app.post("/ocr/url")
+async def ocr_from_url(request: OCRURLRequest):
+    """
+    OCR endpoint that accepts image URL
+    """
+    import requests
+    
+    try:
+        # Download image from URL
+        response = requests.get(request.url, timeout=10)
+        response.raise_for_status()
+        
+        image = Image.open(io.BytesIO(response.content))
+        
+        # Convert to RGB numpy array
+        if image.mode != 'RGB':
+            image = image.convert('RGB')
+        img_np = np.array(image)
+        
+        # Process OCR
+        results = process_ocr(img_np, request.use_det, request.use_cls, request.use_rec)
+        
+        return {
+            "success": True,
+            "texts": [item["text"] for item in results],
+            "details": results,
+            "num_texts": len(results)
+        }
+    
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error processing image from URL: {str(e)}")
+
+# Tạo Gradio interface
+gradio_interface = gr.Interface(
+    fn=recognize_text_gradio,
+    inputs=[
+        gr.Image(label="Upload Image", type="numpy"),
+        gr.Checkbox(label="use_det", value=True),
+        gr.Checkbox(label="use_cls", value=True),
+        gr.Checkbox(label="use_rec", value=True),
+    ],
+    outputs=gr.Textbox(label="OCR Results", lines=10),
+    title="OCR with RapidOCR",
+    description="Upload an image to extract text using RapidOCR"
+)
+
+# Mount Gradio app to FastAPI
+app = gr.mount_gradio_app(fastapi_app, gradio_interface, path="/")
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)