app.py

import os
import tempfile
import base64
import time
import cv2
import streamlit as st
import requests
from ultralytics import YOLO
from huggingface_hub import hf_hub_download
import imageio
import numpy as np

# Page config must be first
st.set_page_config(
    page_title="Wildfire Detection Demo",
    page_icon="🔥",
    layout="wide",
    initial_sidebar_state="expanded"
)

# Helper function to display videos
def show_video(video_bytes: bytes, title: str, loop=True):
    if not video_bytes:
        st.warning(f"No {title} video available.")
        return
    video_base64 = base64.b64encode(video_bytes).decode()
    loop_attr = "loop" if loop else ""
    video_html = f"""
    <h4>{title}</h4>
    <video width="100%" controls autoplay muted {loop_attr}>
      <source src="data:video/mp4;base64,{video_base64}" type="video/mp4">
      Your browser does not support the video tag.
    </video>
    """
    st.markdown(video_html, unsafe_allow_html=True)

# Initialize session state
for key in ["processed_video", "processing_complete", "start_time", "progress"]:
    if key not in st.session_state:
        st.session_state[key] = None if key in ["processed_video", "start_time"] else False if key == "processing_complete" else 0

# Load model
@st.cache_resource
def load_model():
    repo_id = "tstone87/ccr-colorado"
    filename = "best.pt"
    try:
        model_path = hf_hub_download(repo_id=repo_id, filename=filename, repo_type="model")
        return YOLO(model_path)
    except Exception as e:
        st.error(f"Failed to load model: {str(e)}")
        return None

model = load_model()

# Sidebar
with st.sidebar:
    st.header("Process Your Own Video")
    uploaded_file = st.file_uploader("Upload a video", type=["mp4"])
    confidence = st.slider("Detection Confidence", 0.25, 1.0, 0.4)
    fps_options = {
        "Original FPS": None,
        "3 FPS": 3,
        "1 FPS": 1,
        "1 frame/4s": 0.25,
        "1 frame/10s": 0.1,
        "1 frame/15s": 0.0667,
        "1 frame/30s": 0.0333
    }
    selected_fps = st.selectbox("Output FPS", list(fps_options.keys()), index=0)
    process_button = st.button("Process Video")
    progress_bar = st.progress(0)
    progress_text = st.empty()
    download_slot = st.empty()

# Main content
st.title("Wildfire Detection Demo")
st.markdown("Watch our example videos below or upload your own in the sidebar!")

# Example videos
example_videos = {
    "T Example": ("T1.mp4", "T2.mpg"),
    "LA Example": ("LA1.mp4", "LA2.mp4")
}

for example_name in example_videos:
    col1, col2 = st.columns(2)
    orig_file, proc_file = example_videos[example_name]
    try:
        orig_url = f"https://huggingface.co/spaces/tstone87/ccr-colorado/resolve/main/{orig_file}"
        proc_url = f"https://huggingface.co/spaces/tstone87/ccr-colorado/resolve/main/{proc_file}"
        orig_data = requests.get(orig_url).content
        proc_data = requests.get(proc_url).content
        
        with col1:
            show_video(orig_data, f"{example_name} - Original", loop=True)
        with col2:
            show_video(proc_data, f"{example_name} - Processed", loop=True)
    except Exception as e:
        st.error(f"Failed to load {example_name}: {str(e)}")

# Video processing
def process_video(video_file, target_fps, confidence):
    with tempfile.NamedTemporaryFile(suffix=".mp4", delete=False) as tmp:
        tmp.write(video_file.read())
        tmp_path = tmp.name
    
    try:
        reader = imageio.get_reader(tmp_path)
        meta = reader.get_meta_data()
        original_fps = meta['fps']
        width, height = meta['size']
        total_frames = meta['nframes'] if meta['nframes'] != float('inf') else 1000  # Fallback for unknown length
        
        output_fps = fps_options[target_fps] if fps_options[target_fps] else original_fps
        frame_interval = max(1, int(original_fps / output_fps)) if output_fps else 1
        
        out_path = tempfile.NamedTemporaryFile(suffix='.mp4', delete=False).name
        writer = cv2.VideoWriter(out_path, cv2.VideoWriter_fourcc(*'mp4v'), output_fps or original_fps, (width, height))
        
        st.session_state.start_time = time.time()
        processed_count = 0
        
        for i, frame in enumerate(reader):
            if i % frame_interval == 0:
                frame_rgb = np.array(frame)
                results = model.predict(frame_rgb, conf=confidence)
                processed_frame = results[0].plot()[:, :, ::-1]
                writer.write(processed_frame)
                
                processed_count += 1
                elapsed = time.time() - st.session_state.start_time
                progress = (i + 1) / total_frames
                st.session_state.progress = min(progress, 1.0)
                
                if elapsed > 0:
                    frames_left = total_frames - i - 1
                    time_per_frame = elapsed / processed_count
                    eta = frames_left * time_per_frame / frame_interval
                    eta_str = f"{int(eta // 60)}m {int(eta % 60)}s"
                else:
                    eta_str = "Calculating..."
                
                progress_bar.progress(st.session_state.progress)
                progress_text.text(f"Progress: {st.session_state.progress:.1%} | ETA: {eta_str}")
        
        writer.release()
        reader.close()
        
        with open(out_path, 'rb') as f:
            return f.read()
    
    finally:
        if os.path.exists(tmp_path):
            os.unlink(tmp_path)
        if os.path.exists(out_path):
            os.unlink(out_path)

# Process uploaded video
if process_button and uploaded_file and model:
    with st.spinner("Processing video..."):
        st.session_state.processed_video = process_video(uploaded_file, selected_fps, confidence)
        st.session_state.processing_complete = True
        progress_bar.progress(1.0)
        progress_text.text("Processing complete!")

# Show processed video and download button
if st.session_state.processing_complete and st.session_state.processed_video:
    st.subheader("Your Processed Video")
    show_video(st.session_state.processed_video, "Processed Result", loop=False)
    download_slot.download_button(
        label="Download Processed Video",
        data=st.session_state.processed_video,
        file_name="processed_wildfire.mp4",
        mime="video/mp4"
    )

if not model:
    st.error("Model loading failed. Please check the repository and model file availability.")