-AI-Based-Terrain-Classification-Using-Deep-Learning-ResNet50-

It is main ly used to detect live capturing with any camera like drone ,mobile and satellite image . Insert in this and then you will se then , It will show disater identify ,it will also show The Recommendaation Action What u will do in that condition

**#CODE **

import numpy as np import matplotlib.pyplot as plt from tensorflow.keras.applications.resnet50 import ResNet50, preprocess_input, decode_predictions from PIL import Image

Load AI model

model = ResNet50(weights="imagenet")

Terrain / weather keywords

nature_keywords = { "Rocky": ["rock","mountain","cliff","alp","ridge","valley"], "Water": ["sea","ocean","lake","river","lakeside","waterfall","water"], "Forest": ["forest","wood","tree","rainforest","pine"], "Grassland": ["meadow","field","pasture","prairie","park","lawn"], "Sandy": ["sand","beach","sandbar","seashore"], "Desert": ["desert","dune"], "Snow": ["snow","ice","glacier"], "Storm": ["storm","cloud","thundercloud","tornado","cyclone","hurricane"], "Fire": ["fire","volcano","lava","eruption"], "Agriculture": ["hay","barn","rapeseed","field","crop","farm"] }

Disaster detection

def detect_disaster(detected_nature): risks = []

if "Storm" in detected_nature:
    risks.append("Tornado / Thunderstorm")

if "Water" in detected_nature:
    risks.append("Flood")

if "Rocky" in detected_nature and "Water" in detected_nature:
    risks.append("Landslide")

if "Snow" in detected_nature and "Rocky" in detected_nature:
    risks.append("Avalanche")

if "Fire" in detected_nature:
    risks.append("Volcanic Eruption")

return risks

Safety recommendations

def recommend_actions(risks):

actions = []

for r in risks:

    if r == "Tornado / Thunderstorm":
        actions.extend([
            "Seek shelter immediately",
            "Avoid open areas",
            "Monitor weather alerts"
        ])

    if r == "Flood":
        actions.extend([
            "Move to higher ground",
            "Avoid flooded roads"
        ])

    if r == "Landslide":
        actions.extend([
            "Stay away from steep slopes",
            "Evacuate unstable terrain"
        ])

    if r == "Avalanche":
        actions.extend([
            "Avoid snowy mountain slopes",
            "Follow avalanche warnings"
        ])

    if r == "Volcanic Eruption":
        actions.extend([
            "Prepare evacuation routes",
            "Keep emergency supplies ready",
            "Monitor volcanic activity alerts",
            "Avoid valleys and low-lying areas near volcanoes"
        ])

return actions

def detect_nature(image_path):

img = Image.open(image_path).convert("RGB")

plt.imshow(img)
plt.axis("off")
plt.title("Input Image")
plt.show()

img = img.resize((224,224))
img_array = np.array(img)
img_array = np.expand_dims(img_array, axis=0)
img_array = preprocess_input(img_array)

preds = model.predict(img_array)
results = decode_predictions(preds, top=5)[0]

labels = []

print("\nDetected Objects\n")

for r in results:
    label = r[1].lower()
    conf = r[2]*100
    labels.append(label)
    print(label,"-",round(conf,2),"%")

# Detect terrain types
detected_nature = set()

for nature,keywords in nature_keywords.items():
    for label in labels:
        if any(k in label for k in keywords):
            detected_nature.add(nature)

print("\nDetected Terrain / Weather\n")
print(detected_nature)

# Disaster risk
risks = detect_disaster(detected_nature)

print("\nDisaster Risk\n")

if not risks:
    print("No major disaster detected")
else:
    for r in risks:
        print(r)

# Recommended actions
actions = recommend_actions(risks)

print("\nRecommended Actions\n")

if not actions:
    print("No action required")
else:
    for a in actions:
        print("-",a)

Run

detect_nature("IMAGE")