import gradio as gr
import tensorflow as tf
from PIL import Image
import numpy as np
import tensorflow as tf
import keras.backend as K


# Define focal loss
def focal_loss(gamma=2., alpha=0.25):
    def focal_loss_fixed(y_true, y_pred):
        y_pred = tf.clip_by_value(y_pred, K.epsilon(), 1. - K.epsilon())
        cross_entropy = -y_true * tf.math.log(y_pred)
        weight = alpha * y_true * tf.math.pow((1 - y_pred), gamma)
        loss = weight * cross_entropy
        return tf.reduce_sum(loss, axis=-1)
    return focal_loss_fixed


# Load models
vgg16_model = tf.keras.models.load_model(
    "vgg16_best_model.keras"
)

xception_model = tf.keras.models.load_model(
    'xception_best.keras',
    custom_objects={'focal_loss_fixed': focal_loss()}
)


def predict_fire(image):
    img = Image.fromarray(image).convert("RGB")

    # Preprocess for vgg16_model (128x128 input size)
    vgg16_img = img.resize((128, 128))
    vgg16_img_array = np.array(vgg16_img) / 255.0
    vgg16_img_array = np.expand_dims(vgg16_img_array, axis=0)

    # Fire detection using vgg16_model
    fire_pred = vgg16_model.predict(vgg16_img_array)
    fire_status = "Fire Detected" if fire_pred[0][0] > 0.5 else "No Fire Detected"

    # If fire is detected, preprocess for xception_model (224x224 input size)
    if fire_status == "Fire Detected":
        xception_img = img.resize((224, 224))
        xception_img_array = np.array(xception_img) / 255.0
        xception_img_array = np.expand_dims(xception_img_array, axis=0)

        # Severity prediction using xception_model
        severity_pred = xception_model.predict(xception_img_array)
        severity_level = np.argmax(severity_pred[0])
        severity = ["Mild", "Moderate", "Severe"][severity_level]
        
        # Static rule-based recommendations with detailed instructions
        if severity == "Mild":
            recommendation = (
                "Fire detected is mild and manageable. "
                "For the Fire Department: Ensure continuous monitoring of the fire. "
                "Deploy fire trucks and extinguishing equipment if necessary to prevent escalation. "
                "For the Public: Stay alert and stay indoors. Evacuate only if advised by authorities. "
                "Ensure clear access routes for emergency services. "
                "Keep fire safety equipment such as fire extinguishers readily available."
            )
        elif severity == "Moderate":
            recommendation = (
                "Fire detected is moderate and poses a significant risk. "
                "For the Fire Department: Immediate response is needed. "
                "Deploy sufficient fire trucks, helicopters (if possible), and personnel to contain the fire. "
                "Establish firebreaks and coordinate with neighboring departments. "
                "For the Public: Evacuate the area promptly as the fire might spread. "
                "Follow evacuation routes and do not return to the area until authorities deem it safe. "
                "Be cautious of smoke inhalation, and wear protective masks if available."
            )
        else:  # Severe
            recommendation = (
                "Severe fire detected with rapid spread potential. Immediate action is critical. "
                "For the Fire Department: Prioritize evacuation operations. "
                "Deploy all available resources, including specialized teams and air support. "
                "Set up perimeters around the affected area and prevent access. "
                "Coordinate with national agencies for additional resources and backup. "
                "For the Public: Evacuate immediately. Leave all belongings behind and proceed to designated safe zones. "
                "Avoid smoke exposure and keep away from fire zones. Follow all official instructions and do not attempt to return to the area until clearance is given by emergency services. "
                "Remain in contact with local authorities for further updates."
            )

    else:
        severity = "N/A"
        recommendation = (
            "No fire detected. However, always be cautious of any unusual smoke or smells in your environment. "
            "Ensure that fire alarms are functioning, and regularly check fire extinguishers. "
            "Stay prepared by familiarizing yourself with fire evacuation routes and emergency contact numbers."
        )

    return fire_status, severity, recommendation


# Gradio interface
interface = gr.Interface(
    fn=predict_fire,
    inputs=gr.Image(type="numpy", label="Upload Image"),
    outputs=[
        gr.Textbox(label="Fire Status"),
        gr.Textbox(label="Severity Level"),
        gr.Textbox(label="Recommendation")
    ],
    title="Fire Prediction and Severity Classification",
    description="Upload an image to predict fire and its severity level (Mild, Moderate, Severe), and get recommendations.",
)

if __name__ == "__main__":
    interface.launch()