import torch
import torchvision.transforms as transforms
from torchvision import models
from PIL import Image

# ✅ Load Class Names
with open("class_names.txt", "r") as f:
    class_names = [line.strip() for line in f.readlines()]

# ✅ Load Trained Model
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = models.mobilenet_v2(pretrained=False)
model.classifier[1] = torch.nn.Linear(model.classifier[1].in_features, len(class_names))
model.load_state_dict(torch.load("plant_disease_model.pth", map_location=device))
model = model.to(device)
model.eval()

# ✅ Image Transformations (Must match training settings)
transform = transforms.Compose([
    transforms.Resize((128, 128)),
    transforms.ToTensor(),
    transforms.Normalize([0.5], [0.5])
])

# ✅ Function to Make Predictions
def predict_image(image_path):
    image = Image.open(image_path).convert("RGB")
    image = transform(image).unsqueeze(0).to(device)

    with torch.no_grad():
        output = model(image)
        predicted_class = torch.argmax(output, dim=1).item()

    return class_names[predicted_class]

# ✅ Test the model (optional)
if __name__ == "__main__":
    sample_image = "test_image.jpg"  # Replace with an actual image path
    prediction = predict_image(sample_image)
    print(f"Predicted Class: {prediction}")