Insiders Hub

Landuse Classification Convolutional Neural Network

Octaviano Pratama

Octaviano Pratama

0 0
  • 0 Collaborators

This project is about how to classify land use images using Convolutional Neural Network(CNN) . We use land use dataset from UC Merced ...learn more

Artificial Intelligence

Groups
Student Developers for AI, DeepLearning

Overview / Usage

BACKGROUND:
This project is about how to classify land use images using Convolutional Neural Network . Land use data provided by UC Merced. This project is developed by using Python3.6, Tensorflow as a backend and Keras as high level deep learning library. Based on dataset, there are 2100 land use images that categorized into 21 classes, so each category has 100 land use images with dimension 256 x 256 pixel. In this project, we will use 85 data for each class as training data and 5 data for each class as testing data, so total 1785 land use images use as training data with 21 class/label and 105 land use images use as testing data with 21 class/label.

RESULT:
accuracy 10 epochs: 47%
accuracy 50 epochs: 60%
accuracy 100 epochs: ?

Comments (0)