Spaces:
Sleeping
Sleeping
File size: 7,533 Bytes
3c40c7a | 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 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 | """
=============================================================================
Bilinear Interpolation -- FastAPI Microservice Backend
-------------------------------------------------------
A standalone REST API that exposes the core numerical interpolation
algorithms and error metrics as HTTP endpoints.
This service is designed to be consumed by ANY client (Streamlit,
React, CLI, mobile, etc.) -- it has zero UI dependencies.
ENDPOINTS
~~~~~~~~~
GET /health -> {"status": "ok"}
POST /upscale -> upscaled image (base64) + timing
POST /evaluate -> MSE / PSNR / MAE metrics (JSON)
HOW TO RUN
~~~~~~~~~~
pip install fastapi uvicorn python-multipart opencv-python numpy
uvicorn api:app --reload --port 8000
Stack: FastAPI, Uvicorn, NumPy, OpenCV
=============================================================================
"""
import os
import sys
import time
import base64
import io
import numpy as np
import cv2
from fastapi import FastAPI, UploadFile, File, Form, HTTPException
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
# =========================================================================
# Ensure src/ package is importable
# =========================================================================
PROJECT_ROOT = os.path.dirname(os.path.abspath(__file__))
if PROJECT_ROOT not in sys.path:
sys.path.insert(0, PROJECT_ROOT)
from src.bilinear_interpolation import bilinear_interpolation, nearest_neighbor
from src.metrics import calculate_error_metrics
# #########################################################################
# APP INITIALISATION
# #########################################################################
app = FastAPI(
title="Bilinear Interpolation API",
description=(
"A microservice exposing manual Nearest-Neighbour and Bilinear "
"Interpolation algorithms for image upscaling, plus numerical "
"error analysis (MSE, PSNR, MAE)."
),
version="1.0.0",
)
# ---- CORS: allow the Streamlit frontend (and any localhost dev) ----
app.add_middleware(
CORSMiddleware,
allow_origins=["*"], # Replace "*" with the specific Streamlit app URL for stricter security
allow_credentials=False, # Must be False when allow_origins is ["*"]
allow_methods=["*"],
allow_headers=["*"],
)
# #########################################################################
# HELPER FUNCTIONS
# #########################################################################
def _bytes_to_bgr(raw_bytes: bytes) -> np.ndarray:
"""Decode raw image bytes into an OpenCV BGR numpy array."""
arr = np.frombuffer(raw_bytes, dtype=np.uint8)
img = cv2.imdecode(arr, cv2.IMREAD_COLOR)
if img is None:
raise HTTPException(
status_code=400,
detail="Could not decode the uploaded image. "
"Ensure it is a valid PNG/JPG/BMP file.",
)
return img
def _bgr_to_base64_png(img: np.ndarray) -> str:
"""Encode a BGR numpy array as a base64 PNG string."""
success, buf = cv2.imencode(".png", img)
if not success:
raise HTTPException(status_code=500, detail="Failed to encode image.")
return base64.b64encode(buf.tobytes()).decode("utf-8")
def _scale_image(image: np.ndarray, factor: int, algorithm: str):
"""
Upscale *image* by *factor* using the specified *algorithm*.
Returns (result_bgr, elapsed_seconds).
Delegates to the manual loop-based implementations -- NOT cv2.resize.
"""
h, w = image.shape[:2]
new_h, new_w = int(h * factor), int(w * factor)
t0 = time.perf_counter()
if algorithm == "nearest":
result = nearest_neighbor(image, new_h, new_w, disable_tqdm=True)
elif algorithm == "bilinear":
result = bilinear_interpolation(image, new_h, new_w, disable_tqdm=True)
else:
raise HTTPException(
status_code=400,
detail=f"Unknown algorithm '{algorithm}'. Use 'nearest' or 'bilinear'.",
)
elapsed = time.perf_counter() - t0
return result, elapsed
# #########################################################################
# RESPONSE MODELS
# #########################################################################
class UpscaleResponse(BaseModel):
"""Response schema for /upscale."""
image_b64: str
elapsed: float
width: int
height: int
algorithm: str
scale_factor: int
class EvaluateResponse(BaseModel):
"""Response schema for /evaluate."""
mse: float
psnr: float
mae: float
class HealthResponse(BaseModel):
"""Response schema for /health."""
status: str
# #########################################################################
# ENDPOINTS
# #########################################################################
@app.get("/health", response_model=HealthResponse)
async def health_check():
"""Simple liveness probe."""
return {"status": "ok"}
@app.post("/upscale", response_model=UpscaleResponse)
async def upscale_image(
image: UploadFile = File(..., description="Source image file (PNG/JPG)"),
scale_factor: int = Form(4, description="Integer upscale factor (2-8)"),
algorithm: str = Form("bilinear", description="'nearest' or 'bilinear'"),
):
"""
Upscale an uploaded image using the specified algorithm.
Returns the result as a base64-encoded PNG string alongside
timing and dimension metadata.
"""
# ---- Validate inputs ----
if scale_factor < 1 or scale_factor > 16:
raise HTTPException(status_code=400, detail="scale_factor must be 1-16.")
# ---- Decode the uploaded file ----
raw = await image.read()
src_bgr = _bytes_to_bgr(raw)
# ---- Run the interpolation ----
result, elapsed = _scale_image(src_bgr, scale_factor, algorithm)
# ---- Encode result as base64 PNG ----
result_b64 = _bgr_to_base64_png(result)
dst_h, dst_w = result.shape[:2]
return UpscaleResponse(
image_b64=result_b64,
elapsed=round(elapsed, 4),
width=dst_w,
height=dst_h,
algorithm=algorithm,
scale_factor=scale_factor,
)
@app.post("/evaluate", response_model=EvaluateResponse)
async def evaluate_images(
ground_truth: UploadFile = File(..., description="Ground truth HR image"),
upscaled: UploadFile = File(..., description="Upscaled image to evaluate"),
):
"""
Calculate numerical error metrics (MSE, PSNR, MAE) between
a ground truth image and an upscaled reconstruction.
Both images must have the same dimensions.
"""
gt_bytes = await ground_truth.read()
up_bytes = await upscaled.read()
gt_bgr = _bytes_to_bgr(gt_bytes)
up_bgr = _bytes_to_bgr(up_bytes)
if gt_bgr.shape != up_bgr.shape:
raise HTTPException(
status_code=400,
detail=(
f"Shape mismatch: ground_truth={gt_bgr.shape}, "
f"upscaled={up_bgr.shape}. Both must be identical."
),
)
metrics = calculate_error_metrics(gt_bgr, up_bgr)
# Handle inf PSNR for JSON serialisation
psnr_val = metrics["psnr"]
if psnr_val == float('inf'):
psnr_val = 999.0 # sentinel for "perfect match"
return EvaluateResponse(
mse=round(metrics["mse"], 4),
psnr=round(psnr_val, 4),
mae=round(metrics["mae"], 4),
)
|