Upload 5 files

README.md

@@ -1,14 +1,6 @@
----
-title: Smartpack Freshness Detector
-emoji: 🏆
-colorFrom: gray
-colorTo: purple
-sdk: gradio
-sdk_version: 5.49.1
-app_file: app.py
-pinned: false
-license: mit
-short_description: An AI-powered SmartPack system that detects fish freshness a
----
+# SmartPack - Mackerel Freshness Detector (Hugging Face Space)
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Upload an indicator image → get freshness class and TVB-N (demo-level).
+
+NOTE:
+TVB-N regressor uses literature/synthetic data → demonstration-only.

app.py

@@ -0,0 +1,103 @@
+import gradio as gr
+import os, joblib, numpy as np, pandas as pd, cv2
+from PIL import Image
+
+CLASSIFIER_PATH = "models/final_rfc.joblib"
+REGRESSOR_PATH = "models/final_rfr.joblib"
+
+FEATURE_COLS = [
+    'mean_R','mean_G','mean_B',
+    'std_R','std_G','std_B',
+    'mean_H','mean_S','mean_V',
+    'std_H','std_S','std_V',
+    'brightness','ratio_R_G','ratio_R_B'
+]
+
+def autocrop(img_pil):
+    arr = cv2.cvtColor(np.array(img_pil), cv2.COLOR_RGB2BGR)
+    gray = cv2.cvtColor(arr, cv2.COLOR_BGR2GRAY)
+    th = cv2.adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY_INV,21,10)
+    kernel = np.ones((5,5), np.uint8)
+    th = cv2.morphologyEx(th, cv2.MORPH_CLOSE, kernel)
+    cnts,_ = cv2.findContours(th, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
+    if not cnts:
+        return img_pil
+    c = max(cnts, key=cv2.contourArea)
+    x,y,w,h = cv2.boundingRect(c)
+    crop = arr[y:y+h, x:x+w]
+    crop = cv2.cvtColor(crop, cv2.COLOR_BGR2RGB)
+    return Image.fromarray(crop)
+
+def extract_features(pil):
+    arr = np.array(pil.convert('RGB'))
+    mean_rgb = arr.mean(axis=(0,1))
+    std_rgb = arr.std(axis=(0,1))
+    hsv = cv2.cvtColor(arr, cv2.COLOR_RGB2HSV)
+    mean_hsv = hsv.mean(axis=(0,1))
+    std_hsv = hsv.std(axis=(0,1))
+    R,G,B = mean_rgb
+    brightness = 0.299*R + 0.587*G + 0.114*B
+    ratio_R_G = R/(G+1e-6)
+    ratio_R_B = R/(B+1e-6)
+    feats = {
+        'mean_R': float(R), 'mean_G': float(G), 'mean_B': float(B),
+        'std_R': float(std_rgb[0]), 'std_G': float(std_rgb[1]), 'std_B': float(std_rgb[2]),
+        'mean_H': float(mean_hsv[0]), 'mean_S': float(mean_hsv[1]), 'mean_V': float(mean_hsv[2]),
+        'std_H': float(std_hsv[0]), 'std_S': float(std_hsv[1]), 'std_V': float(std_hsv[2]),
+        'brightness': float(brightness),
+        'ratio_R_G': float(ratio_R_G), 'ratio_R_B': float(ratio_R_B)
+    }
+    return feats
+
+clf = None
+rfr = None
+try:
+    if os.path.exists(CLASSIFIER_PATH):
+        clf = joblib.load(CLASSIFIER_PATH)
+except Exception as e:
+    print('Error loading classifier:', e)
+
+try:
+    if os.path.exists(REGRESSOR_PATH):
+        rfr = joblib.load(REGRESSOR_PATH)
+except Exception as e:
+    print('Error loading regressor:', e)
+
+def predict(img):
+    if img is None:
+        return "No image uploaded", None, None
+    cropped = autocrop(img)
+    feats = extract_features(cropped)
+    X = pd.DataFrame([feats])[FEATURE_COLS]
+
+    cls_pred = "Classifier not found"
+    tvb = None
+
+    try:
+        if clf is not None:
+            cls_pred = str(clf.predict(X)[0])
+    except Exception as e:
+        cls_pred = f"Error: {e}"
+
+    try:
+        if rfr is not None:
+            tvb = float(rfr.predict(X)[0])
+    except:
+        tvb = None
+
+    return cls_pred, tvb, cropped
+
+demo = gr.Interface(
+    fn=predict,
+    inputs=gr.Image(type="pil", label="Upload indicator image"),
+    outputs=[
+        gr.Textbox(label="Freshness class"),
+        gr.Number(label="Estimated TVB-N (mg/100g)"),
+        gr.Image(label="Cropped indicator")
+    ],
+    title="SmartPack — Mackerel Freshness Detector",
+    description="Upload an indicator image to classify freshness and estimate amine concentration."
+)
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)

final_rfc.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff7da957060eb606316334da5fcd8ea99cf85f1f5ffc3b08dbd73c65194b6e26
+size 577969

final_rfr.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11d3b924e716d419c863ff7aec6cf88795423deac3efd8997a2978e5c2e01cbf
+size 128625

requirements.txt

@@ -0,0 +1,7 @@
+gradio==3.40.1
+opencv-python-headless
+numpy
+pandas
+joblib
+Pillow
+scikit-learn