This repository contains the implementation of a terrain recognition system using deep learning. The primary goal of this project is to detect, classify, and predict terrain conditions, with a specific focus on roughness and slipperiness. Leveraging the Xception architecture and Convolutional Neural Networks (CNNs), our model aims to enhance terrain understanding in various applications.
-
Terrain Classification: The model can accurately classify diverse terrain types, including Grassy, Rocky, Sandy, and Marshy.
-
Prediction of Terrain Characteristics: In addition to classification, the CNN-based model predicts the roughness or slipperiness of the terrain, providing valuable insights for applications such as autonomous navigation and outdoor activity planning.
-
Efficient Architecture: We have chosen the Xception architecture for its efficiency and effectiveness in processing visual data.
-
Robust Training Dataset: The dataset has been meticulously curated to include a variety of terrain types, ensuring the model's robustness.
- Python
- TensorFlow
- keras
- matplotlib
- sklearn
- seaborn
- numpy
-
Clone the repository:
git clone https://github.com/your-username/terrain-recognition.git cd terrain-recognition
###Install dependencies -pip install -r requirements.txt
-
TensorFlow
-
keras
-
matplotlib
-
sklearn
-
seaborn
-
numpy
###Dataset
#With the help of this link you will access the Dataset
https://www.kaggle.com/datasets/ai21ds06anilriswal/terrain-dataset