NewParser / main.py
Sanket17's picture
Update main.py
9f9625c verified
from fastapi import FastAPI, File, UploadFile, HTTPException
from fastapi.responses import JSONResponse
from pydantic import BaseModel
import base64
import io
import os
from PIL import Image
import torch
import numpy as np
from ultralytics import YOLO
from transformers import AutoProcessor, AutoModelForCausalLM
# Ensure directories exist
if not os.path.exists("weights/icon_detect"):
os.makedirs("weights/icon_detect")
# Model loading with error handling
try:
# Load YOLO model
yolo_model = YOLO("weights/icon_detect/best.pt").to("cuda")
except Exception as e:
print(f"Error loading YOLO model: {e}")
yolo_model = YOLO("weights/icon_detect/best.pt") # Load on CPU if CUDA fails
# Load Caption Model (Florence and OmniParser)
try:
processor = AutoProcessor.from_pretrained("microsoft/Florence-2-base", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
"microsoft/OmniParser",
torch_dtype=torch.float16,
trust_remote_code=True
).to("cuda")
except Exception as e:
print(f"Error loading caption model: {e}")
processor = AutoProcessor.from_pretrained("microsoft/Florence-2-base", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
"microsoft/OmniParser",
torch_dtype=torch.float16,
trust_remote_code=True
)
caption_model_processor = {"processor": processor, "model": model}
print("Finished loading models!")
# FastAPI app initialization
app = FastAPI()
# Pydantic response model
class ProcessResponse(BaseModel):
image: str # Base64 encoded image
parsed_content_list: str
label_coordinates: str
# Function to process the image, apply YOLO, and generate captions
def process(
image_input: Image.Image, box_threshold: float, iou_threshold: float
) -> ProcessResponse:
image_save_path = "imgs/saved_image_demo.png"
image_input.save(image_save_path)
image = Image.open(image_save_path)
# Ratio for bounding box scaling
box_overlay_ratio = image.size[0] / 3200
draw_bbox_config = {
"text_scale": 0.8 * box_overlay_ratio,
"text_thickness": max(int(2 * box_overlay_ratio), 1),
"text_padding": max(int(3 * box_overlay_ratio), 1),
"thickness": max(int(3 * box_overlay_ratio), 1),
}
# OCR Box Detection and Filtering (using EasyOCR and PaddleOCR)
try:
ocr_bbox_rslt, is_goal_filtered = check_ocr_box(
image_save_path,
display_img=False,
output_bb_format="xyxy",
goal_filtering=None,
easyocr_args={"paragraph": False, "text_threshold": 0.9},
use_paddleocr=True,
)
text, ocr_bbox = ocr_bbox_rslt
except Exception as e:
raise HTTPException(status_code=500, detail=f"OCR processing failed: {e}")
# YOLO and Caption Model Inference
try:
dino_labled_img, label_coordinates, parsed_content_list = get_som_labeled_img(
image_save_path,
yolo_model,
BOX_TRESHOLD=box_threshold,
output_coord_in_ratio=True,
ocr_bbox=ocr_bbox,
draw_bbox_config=draw_bbox_config,
caption_model_processor=caption_model_processor,
ocr_text=text,
iou_threshold=iou_threshold,
)
except Exception as e:
raise HTTPException(status_code=500, detail=f"YOLO or caption model inference failed: {e}")
# Convert processed image to base64
image = Image.open(io.BytesIO(base64.b64decode(dino_labled_img)))
parsed_content_list_str = "\n".join(parsed_content_list)
# Encode image to base64
buffered = io.BytesIO()
image.save(buffered, format="PNG")
img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
return ProcessResponse(
image=img_str,
parsed_content_list=str(parsed_content_list_str),
label_coordinates=str(label_coordinates),
)
# FastAPI route to process uploaded image
@app.post("/process_image", response_model=ProcessResponse)
async def process_image(
image_file: UploadFile = File(...),
box_threshold: float = 0.05,
iou_threshold: float = 0.1,
):
try:
contents = await image_file.read()
image_input = Image.open(io.BytesIO(contents)).convert("RGB")
except Exception as e:
raise HTTPException(status_code=400, detail=f"Invalid image file: {e}")
# Process the image
response = process(image_input, box_threshold, iou_threshold)
return response