🌄 Intel Image Classification using CNN (Deep Learning)

A deep learning project focused on building and evaluating a Convolutional Neural Network (CNN) for multi-class image classification using the Intel Image Classification dataset.

---

🚀 Run Notebook in Google Colab

Click the button below to open the notebook directly in Google Colab.

---

📘 Project Overview

This project implements a Convolutional Neural Network (CNN) to classify natural scene images into multiple categories.

The project covers the complete deep learning pipeline including:

• Dataset exploration

• Image preprocessing

• Data augmentation

• CNN architecture design

• Model training and validation

• Model evaluation using performance metrics

• Prediction visualization

The goal is to build a robust CNN model capable of accurately classifying real-world natural scenes.

---

🎯 Objective

The main objectives of this project are:

🔹 Understand the structure of an image dataset

🔹 Perform image preprocessing and augmentation

🔹 Build a CNN architecture from scratch

🔹 Train the model using TensorFlow/Keras

🔹 Evaluate model performance using multiple metrics

🔹 Visualize model predictions on unseen test images

---

📂 Dataset Information

The dataset used is the Intel Image Classification Dataset available on Kaggle.

Dataset Link:

https://www.kaggle.com/datasets/puneet6060/intel-image-classification

Dataset Details

Feature	Description
Total Images	25,000+
Image Type	Natural Scene Images
Number of Classes	6
Image Size	Resized to 150 × 150

Target Classes

The dataset contains 6 scene categories:

• 🏢 Buildings

• 🌲 Forest

• 🧊 Glacier

• ⛰ Mountain

• 🌊 Sea

• 🛣 Street

---

🧹 Data Preprocessing

Before training the CNN model, several preprocessing steps were applied:

✔ Image Resizing

All images were resized to 150 × 150 pixels to maintain uniform input size.

✔ Pixel Normalization

Pixel values were scaled to the range 0–1 by dividing by 255.

✔ Train / Validation / Test Split

The dataset was split into:

• Training Set

• Validation Set

• Test Set

✔ Data Augmentation

To improve model generalization and prevent overfitting, the following augmentation techniques were used:

• Image Rotation

• Horizontal Flip

• Zoom

• Width & Height Shift

• Shear Transform

---

🧠 CNN Model Architecture

A Convolutional Neural Network (CNN) was built from scratch using TensorFlow / Keras.

Architecture Components

✔ Convolutional Layers

✔ ReLU Activation Function

✔ Max Pooling Layers

✔ Dropout Layers for Regularization

✔ Fully Connected (Dense) Layers

✔ Softmax Output Layer for Multi-Class Classification

---

⚙ Model Training

The CNN model was trained using the training dataset and validated using the validation dataset.

Training Techniques Used

• Early Stopping

• Model Checkpointing

• Batch Training

• Validation Monitoring

Training progress was monitored using:

• Training Accuracy vs Validation Accuracy

• Training Loss vs Validation Loss

---

📊 Model Evaluation

After training, the model was evaluated using the test dataset.

Evaluation Metrics

• Accuracy

• Precision

• Recall

• F1 Score

• Confusion Matrix

---

🔍 Prediction Visualization

To better understand the model’s performance:

• 10 random test images were selected

• Each image was displayed with:

  • Actual Label

  • Predicted Label

---

🛠 Tech Stack

Tool	Purpose
Python	Programming language
TensorFlow / Keras	Deep learning framework
NumPy	Numerical computation
Matplotlib	Visualization
Seaborn	Confusion matrix visualization
OpenCV / PIL	Image processing
Kaggle API	Dataset download
Google Colab	Development environment

---

📁 Repository Structure

intel-image-classification-cnn/

│

├── DL_Assignment_1_CNN.ipynb

├── README.md

└── DL Assignment 1 - CNN.pdf

---

🚀 How to Run the Project

1️⃣ Open the Notebook

Click the Google Colab button above.

---

2️⃣ Install Required Libraries

pip install tensorflow matplotlib numpy seaborn kaggle

---

3️⃣ Kaggle API Setup

Upload your kaggle.json file and run:

!mkdir -p ~/.kaggle
!cp kaggle.json ~/.kaggle/
!chmod 600 ~/.kaggle/kaggle.json

Download dataset:

!kaggle datasets download -d puneet6060/intel-image-classification

Unzip dataset:

!unzip intel-image-classification.zip

---

4️⃣ Run the Notebook

Run all cells sequentially to:

• Load dataset

• Preprocess images

• Train CNN model

• Evaluate performance

• Visualize predictions

---

🧠 Key Learning Outcomes

✔ Understanding CNN architecture

✔ Image preprocessing techniques

✔ Data augmentation for deep learning

✔ Training deep learning models

✔ Evaluating classification performance

✔ Visualizing CNN predictions

---

📌 Academic Submission

This repository was created as part of a Deep Learning assignment to demonstrate the design, training, and evaluation of a Convolutional Neural Network (CNN) for real-world image classification.

📌 Future Enhancements

• Implement Transfer Learning (VGG16, ResNet, MobileNet) for better accuracy

• Perform hyperparameter tuning to optimize model performance

• Use larger and more diverse image datasets

• Apply advanced augmentation techniques

• Deploy the model as a web application (Streamlit / Flask)

• Add real-time image prediction feature

📌 Future Enhancements

• Implement Transfer Learning (VGG16, ResNet, MobileNet) for better accuracy

• Perform hyperparameter tuning to optimize model performance

• Use larger and more diverse image datasets

• Apply advanced augmentation techniques

• Deploy the model as a web application (Streamlit / Flask)

• Add real-time image prediction feature

  👤 Author

Name: Laya Mary Joy

Organization: Entri Elevate

Date: March 15, 2026

⭐ Acknowledgment

Thanks to Entri Elevate for guidance and support.