Update api.py

api.py

@@ -3,9 +3,11 @@ import json
 import logging
 import os
 from typing import Optional
+from collections import deque
 
 import cv2
 import numpy as np
+import imagehash
 import uvicorn
 from fastapi import FastAPI, File, Form, UploadFile, HTTPException
 from fastapi.middleware.cors import CORSMiddleware
@@ -16,6 +18,9 @@ from ultralytics import YOLO
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
+# In-memory store for recent image hashes (Simple Deduplication)
+RECENT_HASHES = deque(maxlen=100)
+
 app = FastAPI(title="Arise AI API", version="1.0.0")
 
 # CORS
@@ -62,9 +67,24 @@ async def analyze_endpoint(
         pil_image = pil_image.convert("RGB")
         img_np = np.array(pil_image)
 
+        # --- Enhanced Features ---
+        
+        # 1. Spam Detection (Blur Analysis)
+        # Calculate variance of Laplacian: Low variance = Blurry = Potential Spam/Low Quality
+        gray = cv2.cvtColor(img_np, cv2.COLOR_RGB2GRAY)
+        blur_score = cv2.Laplacian(gray, cv2.CV_64F).var()
+        is_spam = blur_score < 100.0  # Threshold can be tuned
+
+        # 2. Deduplication (Perceptual Hashing)
+        phash = str(imagehash.phash(pil_image))
+        is_duplicate = phash in RECENT_HASHES
+        RECENT_HASHES.append(phash)
+
+        logger.info(f"Image Analysis - Spam(Blur): {is_spam} ({blur_score:.2f}), Duplicate: {is_duplicate}, Hash: {phash}")
+
         # Run Inference
         logger.info("Running YOLO inference...")
-        results = model(img_np, conf=0.15)  # Lower confidence to catch more objects
+        results = model(img_np, conf=0.1)  # Lower confidence to 0.1 to catch more objects
         
         detections = []
         primary_issue = "Unknown"
@@ -90,7 +110,7 @@ async def analyze_endpoint(
 
         # Fallback: Check Description if YOLO fails
         if primary_issue == "Unknown" and description:
-            logger.info(f"YOLO failed, checking description: {description}")
+            logger.info(f"YOLO found no objects, checking description: {description}")
             desc_lower = description.lower()
             keywords = {
                 "pothole": "Pothole",
@@ -119,8 +139,9 @@ async def analyze_endpoint(
 
 
         # Process Image for Overlay (if needed)
-        # We can return the base64 of the plotted image
-        annotated_frame = result.plot()
+        # return the base64 of the plotted image with bounding boxes and labels
+        # Customizing plot: line_width=2 (thinner), font_size=1.0 (smaller) for cleaner mobile view
+        annotated_frame = result.plot(line_width=2, font_size=1.0)
         is_success, buffer = cv2.imencode(".jpg", cv2.cvtColor(annotated_frame, cv2.COLOR_RGB2BGR))
         processed_image_base64 = None
         if is_success:
@@ -160,6 +181,20 @@ async def analyze_endpoint(
             "processed_image": processed_image_base64,
             "resolution_estimation": {
                 "estimated_hours": 24 if severity == "High" else 48
+            },
+            "spam_analysis": {
+                "is_spam": is_spam,
+                "spam_score": round(blur_score, 2),
+                "reason": "Image is too blurry" if is_spam else None
+            },
+            "deduplication": {
+                "is_duplicate": is_duplicate,
+                "image_hash": phash,
+                "method": "phash"
+            },
+            "steganography": {
+                "has_hidden_data": False,
+                "method": None
             }
         }