Land Cover Classification Using CNN

The workflow involves preprocessing satellite images, including normalization and resizing, followed by the extraction of relevant features for classification. A convolutional neural network (CNN) is employed to learn spatial patterns and spectral signatures associated with various land cover classes.

Dataset: http://madm.dfki.de/files/sentinel/EuroSAT.zip

Land Cover Types

0: b'AnnualCrop' 
1: b'Forest' 
2: b'HerbaceousVegetation'
3: b'Highway'
4: b'Industrial'
5: b'Pasture'
6: b'PermanentCrop'
7: b'Residential'
8: b'River'
9: b'SeaLake'

Sample Image

5: b'Pasture'

Model Acuracy and Loss Graph

Test loss: 0.5924096703529358
Test accuracy: 0.7912963032722473

Prediction

1/1 [==============================] - 0s 138ms/step
Predicted Class Index: 7
Predicted Class Name: b'Residential'