Update app.py

app.py

@@ -209,6 +209,35 @@ class ClimatePredictor:
             transforms.Normalize(mean=[0.5], std=[0.5])
         ])
 
+    def highlight_top_potential(self, rgb_image, wind_map, solar_map, percentile=95):
+        """상위 포텐셜 지역을 하이라이트로 표시"""
+        # RGB 이미지를 기본으로 사용
+        result = np.copy(rgb_image)
+        
+        # 히트맵 리사이즈
+        h, w = rgb_image.shape[:2]
+        wind_map_resized = np.array(Image.fromarray((wind_map * 255).astype(np.uint8)).resize((w, h))) / 255.0
+        solar_map_resized = np.array(Image.fromarray((solar_map * 255).astype(np.uint8)).resize((w, h))) / 255.0
+        
+        # 상위 N% 마스크 생성
+        wind_threshold = np.percentile(wind_map_resized, percentile)
+        solar_threshold = np.percentile(solar_map_resized, percentile)
+        
+        wind_mask = wind_map_resized >= wind_threshold
+        solar_mask = solar_map_resized >= solar_threshold
+        
+        # 하이라이트 색상 설정
+        wind_color = np.array([0, 255, 0])  # 녹색
+        solar_color = np.array([255, 0, 0])  # 빨간색
+        
+        # 반투명 오버레이 적용
+        alpha = 0.3
+        result[wind_mask] = result[wind_mask] * (1 - alpha) + wind_color * alpha
+        result[solar_mask] = result[solar_mask] * (1 - alpha) + solar_color * alpha
+        
+        return result.astype(np.uint8)
+
+
     def convert_to_single_channel(self, image_array):
         if len(image_array.shape) == 3:
             return np.dot(image_array[...,:3], [0.2989, 0.5870, 0.1140])
@@ -292,13 +321,13 @@ class ClimatePredictor:
             solar_map = solar_pred.cpu().numpy()[0, 0]
             
             # 결과 시각화
-            fig = plt.figure(figsize=(20, 10))
+            fig = plt.figure(figsize=(20, 12))
             
-            # 1. 원본 이미지와 상위 5% 이모지 오버레이
+            # 1. 원본 이미지와 상위 5% 하이라이트 오버레이
             ax1 = plt.subplot(2, 2, 1)
-            overlay_img = self.overlay_emojis(rgb_image, wind_map, solar_map)
-            ax1.imshow(overlay_img)
-            ax1.set_title('Top 5% Potential Sites')
+            highlighted_img = self.highlight_top_potential(rgb_image, wind_map, solar_map)
+            ax1.imshow(highlighted_img)
+            ax1.set_title('Top 5% Potential Sites\n(Red: Solar, Green: Wind)', pad=20)
             ax1.axis('off')
             
             # 2. 풍력 발전 잠재량 히트맵
@@ -313,13 +342,13 @@ class ClimatePredictor:
                                       cbar_kws={'label': 'Solar Power Potential'})
             ax3.set_title('Solar Power Potential Map')
             
-            # 4. 상위 5% 포텐셜 지역 표시
+            # 4. 상위 5% 포텐셜만 표시한 히트맵
             ax4 = plt.subplot(2, 2, 4)
-            wind_top = np.where(get_top_percentile_mask(wind_map), wind_map, 0)
-            solar_top = np.where(get_top_percentile_mask(solar_map), solar_map, 0)
+            wind_top = np.where(wind_map >= np.percentile(wind_map, 95), wind_map, 0)
+            solar_top = np.where(solar_map >= np.percentile(solar_map, 95), solar_map, 0)
             combined_map = np.stack([solar_top, wind_top, np.zeros_like(wind_map)], axis=-1)
             ax4.imshow(combined_map)
-            ax4.set_title('Top 5% Potential Sites (Red: Solar, Green: Wind)')
+            ax4.set_title('Top 5% Potential Sites Heatmap\n(Red: Solar, Green: Wind)', pad=20)
             ax4.axis('off')
             
             plt.tight_layout()
@@ -375,18 +404,17 @@ def load_examples_from_directory(base_dir):
 
 def create_gradio_interface():
     predictor = ClimatePredictor('best_model.pth')
-
+    
     def process_elevation_file(file_obj):
-        if isinstance(file_obj, str):  # 파일 경로인 경우
+        if isinstance(file_obj, str):
             return np.load(file_obj)
-        else:  # UploadedFile 객체인 경우
+        else:
             return np.load(file_obj.name)
 
     def predict_with_processing(*args):
         rgb_image, ndvi_image, terrain_image, elevation_file = args[:4]
         weather_params = args[4:]
         
-        # elevation 파일 처리
         elevation_data = process_elevation_file(elevation_file)
         
         return predictor.predict_from_inputs(
@@ -395,36 +423,30 @@ def create_gradio_interface():
         )
     
     with gr.Blocks() as interface:
-        gr.Markdown("# Renewable Energy Potential Predictor")
-        gr.Markdown("Upload satellite imagery and environmental data to predict wind and solar power potential.")
-        
-        with gr.Row():
-            # 입력 섹션 (1/3 크기)
-            with gr.Column(scale=1):
-                rgb_input = gr.Image(label="RGB Satellite Image", type="numpy", height=200)
-                ndvi_input = gr.Image(label="NDVI Image", type="numpy", height=200)
-                terrain_input = gr.Image(label="Terrain Map", type="numpy", height=200)
-                elevation_input = gr.File(label="Elevation Data (NPY file)", height=50)
+        with gr.Column():
+            gr.Markdown("# Renewable Energy Potential Predictor")
+            
+            # 입력 섹션 - 모든 입력을 하나의 Row에 배치
+            with gr.Row(equal_height=True):
+                # 이미지/파일 입력
+                rgb_input = gr.Image(label="RGB Satellite Image", type="numpy", height=150, scale=1)
+                ndvi_input = gr.Image(label="NDVI Image", type="numpy", height=150, scale=1)
+                terrain_input = gr.Image(label="Terrain Map", type="numpy", height=150, scale=1)
+                elevation_input = gr.File(label="Elevation Data (NPY)", height=150, scale=1)
                 
-                with gr.Row():
+                # 날씨 파라미터 입력
+                with gr.Column(scale=1):
                     wind_speed = gr.Number(label="Wind Speed (m/s)", value=5.0)
                     wind_direction = gr.Number(label="Wind Direction (°)", value=180.0)
-                
-                with gr.Row():
                     temperature = gr.Number(label="Temperature (°C)", value=25.0)
                     humidity = gr.Number(label="Humidity (%)", value=60.0)
-                
-                predict_btn = gr.Button("Generate Predictions", variant="primary")
+                    predict_btn = gr.Button("Generate Predictions", variant="primary", size="lg")
             
-            # 출력 섹션 (2/3 크기)
-            with gr.Column(scale=2):
-                output_plot = gr.Plot(label="Prediction Results")
-
-        # 예제 데이터 로드
-        examples = load_examples_from_directory("filtered_climate_data")
-        
-        # 예제 갤러리 생성
-        with gr.Row():
+            # 출력 섹션 - 전체 너비 사용
+            output_plot = gr.Plot(label="Prediction Results", container=True, height=600)
+            
+            # 예제 섹션
+            examples = load_examples_from_directory("filtered_climate_data")
             gr.Examples(
                 examples=examples,
                 inputs=[rgb_input, ndvi_input, terrain_input, elevation_input,
@@ -436,7 +458,6 @@ def create_gradio_interface():
                 examples_per_page=5
             )
         
-        # 예측 버튼 클릭 이벤트 연결
         predict_btn.click(
             fn=predict_with_processing,
             inputs=[rgb_input, ndvi_input, terrain_input, elevation_input,