models/rice_leaf_validator.py

import torch
import torch.nn as nn
import torch.nn.functional as F
from torchvision import transforms
from PIL import Image

class RiceLeafValidator(nn.Module):
    """
    CNN model for rice leaf validation using BINARY CLASSIFICATION.
    Trained on two classes: rice_leaf (0) and not_rice_leaf (1).
    """
    def __init__(self):
        super(RiceLeafValidator, self).__init__()
        # Convolutional layers
        self.conv1 = nn.Conv2d(3, 16, kernel_size=3, padding=1)
        self.conv2 = nn.Conv2d(16, 32, kernel_size=3, padding=1)
        self.conv3 = nn.Conv2d(32, 64, kernel_size=3, padding=1)
        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)

        # Fully connected layers
        self.fc1 = nn.Linear(64 * 28 * 28, 128)
        self.fc2 = nn.Linear(128, 2)  # 2 classes: rice_leaf and not_rice_leaf
        self.dropout = nn.Dropout(0.5)

    def forward(self, x):
        # Convolutional layers with pooling
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = self.pool(F.relu(self.conv3(x)))

        # Flatten and fully connected layers
        x = x.view(-1, 64 * 28 * 28)
        x = F.relu(self.fc1(x))
        x = self.dropout(x)
        x = self.fc2(x)

        return x
# Preprocessing pipeline
data_transforms = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

def is_rice_leaf(image_path, model, device, threshold=0.5):
    """
    Determines if an image contains a rice leaf using BINARY classification.

    Args:
        image_path: Path to the image file
        model: Trained binary RiceLeafValidator model
        device: torch device (cpu or cuda)
        threshold: Probability threshold for rice_leaf class (default: 0.5)

    Returns:
        (is_rice_leaf, confidence): Tuple of boolean result and confidence score (0-1)
    """
    model.eval()
    image = Image.open(image_path).convert('RGB')
    image_tensor = data_transforms(image).unsqueeze(0).to(device)

    with torch.no_grad():
        outputs = model(image_tensor)
        probabilities = F.softmax(outputs, dim=1)
        rice_leaf_prob = probabilities[0][0].item()  # Probability for rice_leaf class (index 0)

    is_rice = rice_leaf_prob >= threshold

    return is_rice, rice_leaf_prob