CLAHE

app.py

@@ -2,42 +2,38 @@ import gradio as gr
 import cv2, numpy as np
 from PIL import Image
 
-def uniform_lighting(image):
-    # Convert PIL → OpenCV
+def clahe_uniform(image):
+    # Convert PIL → OpenCV BGR
     img = cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)
 
-    # 1️⃣ Convert to grayscale to estimate illumination
-    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
+    # --- Convert to LAB color space ---
+    lab = cv2.cvtColor(img, cv2.COLOR_BGR2LAB)
+    L, A, B = cv2.split(lab)
 
-    # 2️⃣ Smooth illumination field (large Gaussian kernel)
-    illumination = cv2.GaussianBlur(gray, (51, 51), 0)
+    # --- Apply CLAHE (Contrast Limited Adaptive Histogram Equalization) ---
+    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))
+    L2 = clahe.apply(L)
 
-    # 3️⃣ Compute normalization map using brightest region
-    brightest = np.max(illumination)
-    illum_norm = brightest / (illumination + 1e-5)
-    illum_norm = np.repeat(illum_norm[:, :, np.newaxis], 3, axis=2)
+    # --- Merge back channels and convert to RGB ---
+    merged = cv2.merge([L2, A, B])
+    result = cv2.cvtColor(merged, cv2.COLOR_LAB2BGR)
+    result_rgb = cv2.cvtColor(result, cv2.COLOR_BGR2RGB)
 
-    # 4️⃣ Apply correction
-    corrected = np.clip(img.astype(np.float32) * illum_norm, 0, 255).astype(np.uint8)
+    return [image, Image.fromarray(result_rgb)]
 
-    # Convert back to PIL for Gradio display
-    corrected_rgb = cv2.cvtColor(corrected, cv2.COLOR_BGR2RGB)
-    return Image.fromarray(corrected_rgb)
-
-title = "🧵 Saree AI – Uniform Lighting Correction"
+title = "🧵 Saree AI – Uniform Lighting (CLAHE)"
 description = """
-Upload a saree image with uneven lighting.
-This tool will automatically detect illumination gradients and normalize brightness
-using the brightest region as reference, giving you a flat, uniform look for catalog or dataset use.
+Upload a saree image with uneven illumination.
+This app applies **CLAHE (adaptive histogram equalization)** to balance lighting locally
+while preserving texture and color richness.
 """
 
 demo = gr.Interface(
-    fn=uniform_lighting,
+    fn=clahe_uniform,
     inputs=gr.Image(label="Upload Saree Image", type="pil"),
-    outputs=gr.Image(label="Corrected Image"),
+    outputs=[gr.Image(label="Original"), gr.Image(label="CLAHE Corrected")],
     title=title,
     description=description,
-    examples=[],
 )
 
 if __name__ == "__main__":