https://github.com/arif-miad/plant-species-classification-using-cnn
https://github.com/arif-miad/plant-species-classification-using-cnn
cnn deep-learning image-classification keras python seaborn tensorflow
Last synced: 5 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/arif-miad/plant-species-classification-using-cnn
- Owner: Arif-miad
- License: gpl-3.0
- Created: 2025-01-31T13:38:59.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-01-31T13:46:54.000Z (over 1 year ago)
- Last Synced: 2025-03-27T06:16:11.460Z (over 1 year ago)
- Topics: cnn, deep-learning, image-classification, keras, python, seaborn, tensorflow
- Homepage:
- Size: 280 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
#### 🌱 **Plant Species Classification using CNN**
---
## 📖 **Overview**
This project focuses on **classifying plant species** using a **Convolutional Neural Network (CNN)**. The dataset consists of synthetic plant images categorized into five classes:
- 🌵 **Cactus**
- 🌿 **Fern**
- 🌹 **Rose**
- 🌻 **Sunflower**
- 🌷 **Tulip**
We use **TensorFlow & Keras** to train a deep learning model for **image classification**. The final trained model is evaluated, saved, and can be deployed for real-time predictions.
---
## 📂 **Dataset Structure**
```
plants-classification/
│── dataset/
│ ├── train/
│ │ ├── cactus/
│ │ ├── fern/
│ │ ├── rose/
│ │ ├── sunflower/
│ │ ├── tulip/
│ ├── val/
│ │ ├── cactus/
│ │ ├── fern/
│ │ ├── rose/
│ │ ├── sunflower/
│ │ ├── tulip/
│ ├── train.cache
│ ├── val.cache
```
---
## 🚀 **Project Implementation**
### **✔ 1️⃣ Data Loading**
We load images from the dataset using `tf.keras.preprocessing.image_dataset_from_directory()`.
```python
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
"dataset/train", image_size=(224, 224), batch_size=32
)
val_ds = tf.keras.preprocessing.image_dataset_from_directory(
"dataset/val", image_size=(224, 224), batch_size=32
)
```
---
### **✔ 2️⃣ Preprocessing & Data Augmentation**
We normalize pixel values and apply transformations like flipping and rotation.
```python
data_augmentation = tf.keras.Sequential([
tf.keras.layers.RandomFlip("horizontal"),
tf.keras.layers.RandomRotation(0.2),
])
```
---
### **✔ 3️⃣ CNN Model Creation**
A deep learning model is built using Convolutional Neural Networks (CNNs).
```python
model = tf.keras.Sequential([
tf.keras.layers.Rescaling(1./255, input_shape=(224, 224, 3)),
data_augmentation,
tf.keras.layers.Conv2D(32, (3,3), activation="relu"),
tf.keras.layers.MaxPooling2D(),
tf.keras.layers.Conv2D(64, (3,3), activation="relu"),
tf.keras.layers.MaxPooling2D(),
tf.keras.layers.Conv2D(128, (3,3), activation="relu"),
tf.keras.layers.MaxPooling2D(),
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(128, activation="relu"),
tf.keras.layers.Dropout(0.5),
tf.keras.layers.Dense(5, activation="softmax")
])
```
---
### **✔ 4️⃣ Model Training & Evaluation**
The model is compiled and trained using **Adam optimizer** and **sparse categorical crossentropy loss**.
```python
model.compile(optimizer="adam",
loss="sparse_categorical_crossentropy",
metrics=["accuracy"])
history = model.fit(train_ds, validation_data=val_ds, epochs=20)
```
We then evaluate the trained model:
```python
loss, accuracy = model.evaluate(val_ds)
print(f"Validation Accuracy: {accuracy*100:.2f}%")
```
---
### **✔ 5️⃣ Predictions on New Images**
We load an image and use the trained model to predict its class.
```python
def predict_image(image_path, model, class_names):
img = tf.keras.preprocessing.image.load_img(image_path, target_size=(224, 224))
img_array = tf.keras.preprocessing.image.img_to_array(img)
img_array = tf.expand_dims(img_array, 0)
predictions = model.predict(img_array)
predicted_class = class_names[np.argmax(predictions)]
confidence = np.max(predictions)
plt.imshow(img)
plt.title(f"Prediction: {predicted_class} ({confidence:.2f})")
plt.axis("off")
plt.show()
predict_image("dataset/val/rose/rose1.jpg", model, train_ds.class_names)
```
---
### **✔ 6️⃣ Model Saving & Deployment**
Save the trained model for future use.
```python
model.save("plant_classification_model.h5")
```
Convert to **TensorFlow Lite** for mobile or edge deployment.
```python
converter = tf.lite.TFLiteConverter.from_keras_model(model)
tflite_model = converter.convert()
with open("plant_classification_model.tflite", "wb") as f:
f.write(tflite_model)
```
---
## 📌 **Results & Accuracy**
The CNN model achieves high accuracy in classifying plant species.
📈 **Validation Accuracy:** **~95%**
---
## 🛠 **Technologies Used**
- **Python**
- **TensorFlow / Keras**
- **Matplotlib**
- **NumPy**
- **OpenCV** (for image processing)
---
## 📝 **Future Improvements**
🔹 Use a **larger dataset** for better accuracy.
🔹 Implement **Transfer Learning (ResNet, MobileNet, EfficientNet)** for improved performance.
🔹 Deploy as a **Flask API or Streamlit App** for real-world use.
---
## 🎯 **Conclusion**
This project demonstrates a **CNN-based deep learning model** for classifying plants into different species. The trained model can be further **optimized, fine-tuned, and deployed** for real-world applications.
---
## ⭐ **Contribute**
If you find this project useful, feel free to **⭐ Star the repository** and **Fork it**! 🚀