| import gradio as gr |
| import tensorflow as tf |
| import pandas as pd |
| import numpy as np |
| from datetime import datetime |
| from PIL import Image |
|
|
| MODEL_PATH = "best_model.h5" |
| class_names = ['layak', 'rusak'] |
| history_data = [] |
|
|
| try: |
| best_model = tf.keras.models.load_model(MODEL_PATH) |
| print("✅ Model berhasil dimuat.") |
| except Exception as e: |
| print(f"❌ Gagal memuat model: {e}") |
| |
| best_model = None |
|
|
| |
| def get_detailed_info(label, confidence): |
| if label == 'layak': |
| if confidence > 0.85: |
| return "### ✅ STATUS: SANGAT LAYAK\n**Analisis:** Bangunan dalam kondisi prima. Struktur utama terlihat utuh dan sangat aman untuk dihuni." |
| return "### ⚠️ STATUS: LAYAK (DENGAN CATATAN)\n**Analisis:** Bangunan aman dihuni, namun ditemukan indikasi kerusakan minor. Disarankan pengecekan rutin pada area retakan." |
| else: |
| if confidence > 0.85: |
| return "### 🚨 STATUS: RUSAK PARAH\n**Analisis:** BAHAYA! Ditemukan kerusakan struktur fatal. Segera kosongkan area dan hubungi pihak berwenang." |
| return "### 🚧 STATUS: RUSAK RINGAN\n**Analisis:** Terdeteksi kerusakan fisik pada beberapa bagian. Perlu perbaikan teknis sebelum bangunan dinyatakan aman sepenuhnya." |
|
|
| def predict_image(img): |
| if best_model is None: |
| return {}, "Model belum dimuat." |
| |
| |
| img = img.resize((224, 224)) |
| img_array = tf.keras.preprocessing.image.img_to_array(img) |
| img_array = tf.expand_dims(img_array, 0) / 255.0 |
|
|
| |
| predictions = best_model.predict(img_array)[0] |
| result = {class_names[i]: float(predictions[i]) for i in range(len(class_names))} |
|
|
| top_label = max(result, key=result.get) |
| description = get_detailed_info(top_label, result[top_label]) |
| return result, description |
|
|
| |
| def handle_upload(img): |
| if img is None: return {}, "_Menunggu foto bangunan..._" |
| return predict_image(img) |
|
|
| def handle_report(img, location): |
| if img is None: |
| return {}, pd.DataFrame(history_data, columns=["Waktu", "Status", "Lokasi"]), None, "❌ Gagal: Foto kosong." |
|
|
| output_dict, desc = predict_image(img) |
| status = max(output_dict, key=output_dict.get).upper() |
| now = datetime.now().strftime("%H:%M | %d-%m-%Y") |
|
|
| history_data.insert(0, [now, status, location if location else "Pusat Kota"]) |
| df = pd.DataFrame(history_data, columns=["Waktu", "Status", "Lokasi"]) |
|
|
| return output_dict, df, None, "✅ Laporan berhasil disimpan ke riwayat!" |
|
|
| |
| with gr.Blocks(title="HomeCheck AI") as demo: |
| |
| with gr.Row(): |
| with gr.Column(scale=8): |
| gr.Markdown("# 🏠 HomeCheck AI") |
| gr.Markdown("### *Sistem Deteksi Kelayakan Bangunan Cerdas*") |
| with gr.Column(scale=2): |
| gr.Markdown("") |
|
|
| gr.Markdown("---") |
|
|
| with gr.Tabs(): |
| with gr.TabItem("🔍 Analisis Baru"): |
| with gr.Row(): |
| |
| with gr.Column(variant="panel"): |
| gr.Markdown("#### 📥 Input Data") |
| input_img = gr.Image(sources=["upload", "webcam"], type="pil", label="Foto Bangunan") |
| input_loc = gr.Textbox( |
| label="Titik Lokasi", |
| placeholder="Contoh: Perumahan Indah Blok A, Medan", |
| lines=1 |
| ) |
| btn_report = gr.Button("🚀 SIMPAN LAPORAN", variant="primary") |
|
|
| |
| with gr.Column(): |
| gr.Markdown("#### 📊 Hasil Diagnosa") |
| output_label = gr.Label(num_top_classes=2, label="Probabilitas Akurasi") |
|
|
| with gr.Group(): |
| output_description = gr.Markdown( |
| "**Instruksi:**\nSilakan ambil atau upload foto bagian bangunan yang ingin diperiksa.", |
| ) |
|
|
| with gr.TabItem("📜 Riwayat Pemeriksaan"): |
| gr.Markdown("#### 📑 Log Laporan Tersimpan") |
| output_history = gr.Dataframe( |
| headers=["Waktu", "Status", "Lokasi"], |
| datatype=["str", "str", "str"], |
| interactive=False |
| ) |
|
|
| gr.Markdown("---") |
| gr.Markdown("© 2026 HomeCheck AI") |
|
|
| |
| input_img.change(fn=handle_upload, inputs=input_img, outputs=[output_label, output_description]) |
|
|
| btn_report.click( |
| fn=handle_report, |
| inputs=[input_img, input_loc], |
| outputs=[output_label, output_history, input_img, output_description] |
| ) |
|
|
| if __name__ == "__main__": |
| demo.launch(server_name="0.0.0.0", server_port=7860) |
