Update app.py

app.py

@@ -1,76 +1,70 @@
 import gradio as gr
-from PIL import Image, ImageEnhance, ImageOps, ImageFilter
+import cv2
 import numpy as np
 
-def apply_filter(image, filter_name, intensity):
-    if filter_name == "Soft Glow":
-        glow = image.filter(ImageFilter.GaussianBlur(intensity))
-        enhanced_image = Image.blend(image, glow, 0.5)
-        return enhanced_image
-
-    elif filter_name == "Portrait Enhancer":
-        enhancer = ImageEnhance.Sharpness(image)
-        image = enhancer.enhance(1 + 0.1 * intensity)
-        enhancer = ImageEnhance.Color(image)
-        image = enhancer.enhance(1 + 0.1 * intensity)
-        return image
+def apply_filter(image, filter_type, intensity):
+    # 강도를 0.0에서 1.0 사이로 정규화
+    normalized_intensity = intensity / 100.0
 
-    elif filter_name == "Warm Tone":
-        warm_tone = ImageEnhance.Color(image).enhance(1 + 0.1 * intensity)
-        return warm_tone
-
-    elif filter_name == "Cold Tone":
-        blue_overlay = Image.new('RGB', image.size, (173, 216, 230))
-        blended_image = Image.blend(image, blue_overlay, 0.05 * intensity)
-        enhancer = ImageEnhance.Sharpness(blended_image)
-        sharpened_image = enhancer.enhance(1 + 0.3 * intensity)
-        return sharpened_image
-
-    elif filter_name == "High-Key":
-        enhancer = ImageEnhance.Brightness(image)
-        high_key = enhancer.enhance(1 + 0.1 * intensity)
-        enhancer = ImageEnhance.Contrast(high_key)
-        high_key = enhancer.enhance(0.8 + 0.02 * intensity)
+    if filter_type == "Grayscale":
+        return convert_to_grayscale(image)
+    elif filter_type == "Soft Glow":
+        gaussian = cv2.GaussianBlur(image, (15, 15), 0)
+        soft_glow = cv2.addWeighted(image, 1 - normalized_intensity, gaussian, normalized_intensity, 0)
+        return soft_glow
+    elif filter_type == "Portrait Enhancer":
+        # 참조 코드에 따라 PIL을 사용하여 피부 톤을 균일하게 하고 선명도를 조절
+        from PIL import Image, ImageEnhance
+        image_pil = Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
+        
+        enhancer = ImageEnhance.Sharpness(image_pil)
+        image_pil = enhancer.enhance(1 + 0.1 * intensity)
+        
+        enhancer = ImageEnhance.Color(image_pil)
+        image_pil = enhancer.enhance(1 + 0.1 * intensity)
+        
+        enhanced_image = cv2.cvtColor(np.array(image_pil), cv2.COLOR_RGB2BGR)
+        return enhanced_image
+    elif filter_type == "Warm Tone":
+        increase_red = np.array([[1.0 + 0.2 * normalized_intensity, 0.0, 0.0],
+                                 [0.0, 1.0, 0.0],
+                                 [0.0, 0.0, 1.0 - 0.2 * normalized_intensity]])
+        warm_image = cv2.transform(image, increase_red)
+        return warm_image
+    elif filter_type == "Cold Tone":
+        increase_blue = np.array([[1.0 - 0.2 * normalized_intensity, 0.0, 0.0],
+                                  [0.0, 1.0, 0.0],
+                                  [0.0, 0.0, 1.0 + 0.2 * normalized_intensity]])
+        cold_image = cv2.transform(image, increase_blue)
+        return cold_image
+    elif filter_type == "High-Key":
+        high_key = cv2.convertScaleAbs(image, alpha=1.0 + 0.2 * normalized_intensity, beta=30)
         return high_key
-
-    elif filter_name == "Low-Key":
-        enhancer = ImageEnhance.Contrast(image)
-        low_key = enhancer.enhance(1 + 0.1 * intensity)
-        enhancer = ImageEnhance.Brightness(low_key)
-        low_key = enhancer.enhance(1 - 0.05 * intensity)
+    elif filter_type == "Low-Key":
+        low_key = cv2.convertScaleAbs(image, alpha=1.0 - 0.3 * normalized_intensity, beta=-30)
         return low_key
-
-    elif filter_name == "Haze":
-        haze = image.filter(ImageFilter.GaussianBlur(2 * intensity))
-        enhancer = ImageEnhance.Brightness(haze)
-        haze = enhancer.enhance(1 + 0.05 * intensity)
+    elif filter_type == "Haze":
+        haze = cv2.addWeighted(image, 1.0 - 0.3 * normalized_intensity, np.full(image.shape, 255, dtype=np.uint8), 0.3 * normalized_intensity, 0)
         return haze
-
-    elif filter_name == "Monochrome":
-        mono = image.convert("L")
-        mono = ImageOps.colorize(mono, black="black", white="white")
-        enhancer = ImageEnhance.Contrast(mono)
-        mono = enhancer.enhance(1 + 0.1 * intensity)
-        return mono
-
-    return image
+    else:
+        return image
 
 def convert_to_grayscale(image):
-    gray_image = ImageOps.grayscale(image)
+    gray_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
     return gray_image
 
 def convert_and_save(image, filter_type, intensity):
     filtered_image = apply_filter(image, filter_type, intensity)
     output_path = "output.jpg"
-    filtered_image.save(output_path)
+    cv2.imwrite(output_path, filtered_image)
     return filtered_image, output_path
 
 iface = gr.Interface(
     fn=convert_and_save, 
     inputs=[
-        gr.Image(type="pil", label="이미지 업로드"), 
+        "image", 
         gr.Radio(
-            ["Grayscale", "Soft Glow", "Portrait Enhancer", "Warm Tone", "Cold Tone", "High-Key", "Low-Key", "Haze", "Monochrome"], 
+            ["Grayscale", "Soft Glow", "Portrait Enhancer", "Warm Tone", "Cold Tone", "High-Key", "Low-Key", "Haze"], 
             label="필터 선택"
         ),
         gr.Slider(minimum=1, maximum=100, value=50, label="필터 강도")