Update app.py

app.py

@@ -7,12 +7,11 @@ from ultralytics import YOLO
 from pathlib import Path
 
 # === Konfigurasi ===
-model_path = "best.pt"
+model_path = "best.pt"  # Ganti dengan path model kamu
 class_names = ['coral or rock', 'pipeline', 'ripple marks', 'shipwreck']
-
 model = YOLO(model_path)
 
-# === Fungsi untuk menggambar mask hasil deteksi ===
+# === Fungsi untuk menggambar prediksi lengkap ===
 def draw_predictions(img, segments, class_ids, confs, color=(255, 0, 0), alpha=0.5):
     overlay = img.copy()
     for i, seg in enumerate(segments):
@@ -21,31 +20,40 @@ def draw_predictions(img, segments, class_ids, confs, color=(255, 0, 0), alpha=0
         polygon[:, 1] *= img.shape[0]
         polygon = polygon.astype(np.int32)
 
-        label = f"{class_names[int(class_ids[i])]} ({confs[i]:.2f})"
+        cls_id = int(class_ids[i])  # pastikan integer
+        conf = confs[i]
+
+        # Draw mask
         cv2.fillPoly(overlay, [polygon], color)
+
+        # Draw bounding box
         x, y, w, h = cv2.boundingRect(polygon)
-        cv2.putText(overlay, label, (x, y - 5), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255,255,255), 1)
-        cv2.rectangle(overlay, (x, y), (x + w, y + h), color, 1)
+        cv2.rectangle(overlay, (x, y), (x + w, y + h), color, 2)
+
+        # Draw label + confidence
+        label = f"{class_names[cls_id]} ({conf:.2f})"
+        cv2.putText(overlay, label, (x, y - 8), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255,255,255), 2)
 
     return cv2.addWeighted(overlay, alpha, img, 1 - alpha, 0)
 
-# === Fungsi utama proses dan simpan report ===
-def process_image_and_generate_report(image):
+# === Fungsi proses gambar dan hasil deteksi ===
+def process_image_and_report(image):
     temp_path = "temp.jpg"
     image.save(temp_path)
     img = cv2.imread(temp_path)
     img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
 
-    # Prediksi dari YOLOv8
+    # === Prediksi menggunakan YOLOv8
     results = model(temp_path)[0]
     segs = [seg.xy for seg in results.masks] if results.masks else []
     cls_ids = results.boxes.cls.tolist() if results.boxes else []
     confs = results.boxes.conf.tolist() if results.boxes else []
     xywh = results.boxes.xywhn.tolist() if results.boxes else []
 
-    img_result = draw_predictions(img.copy(), segs, cls_ids, confs)
+    # === Visualisasi prediksi
+    img_pred = draw_predictions(img.copy(), segs, cls_ids, confs)
 
-    # Simpan CSV
+    # === Buat laporan CSV
     rows = []
     for i in range(len(cls_ids)):
         rows.append({
@@ -59,27 +67,25 @@ def process_image_and_generate_report(image):
         })
 
     df = pd.DataFrame(rows)
-    csv_path = "detection_result.csv"
+    csv_path = "detection_report.csv"
     df.to_csv(csv_path, index=False)
 
-    return img_result, csv_path
+    return img_pred, csv_path
 
-# === UI Gradio ===
+# === Gradio UI ===
 with gr.Blocks() as demo:
-    gr.Markdown("## 🎯 YOLOv8 Segmentasi: Prediksi dan Laporan Otomatis")
-    gr.Markdown("Upload gambar untuk mendeteksi objek dan mengunduh hasil deteksi sebagai CSV")
+    gr.Markdown("## 🧠 YOLOv8 Segmentation Viewer + Detection Report")
+    gr.Markdown("Upload image → lihat hasil deteksi (mask, box, label, confidence) → unduh CSV")
 
     with gr.Row():
         with gr.Column():
             image_input = gr.Image(type="pil", label="Upload Image")
-            run_btn = gr.Button("Deteksi dan Simpan Report")
+            submit_btn = gr.Button("Predict and Generate Report")
         with gr.Column():
-            result_image = gr.Image(type="numpy", label="Hasil Prediksi")
-            download_btn = gr.File(label="Unduh CSV Deteksi")
+            image_output = gr.Image(type="numpy", label="Prediction Result")
+            download_csv = gr.File(label="Download Detection CSV")
 
-    run_btn.click(fn=process_image_and_generate_report,
-                  inputs=image_input,
-                  outputs=[result_image, download_btn])
+    submit_btn.click(fn=process_image_and_report, inputs=image_input, outputs=[image_output, download_csv])
 
 if __name__ == "__main__":
     demo.launch()