PratyushTripathy

import os
from pyrsgis import raster

os.chdir("E:\\yourDirectoryName")
mxBangalore = 'l5_Bangalore2011_raw.tif'
builtupBangalore = 'l5_Bangalore2011_builtup.tif'
mxHyderabad = 'l5_Hyderabad2011_raw.tif'

# Read the rasters as array
ds1, featuresBangalore = raster.read(mxBangalore, bands='all')

PratyushTripathy

print("Bangalore multispectral image shape: ", featuresBangalore.shape)
print("Bangalore binary built-up image shape: ", labelBangalore.shape)
print("Hyderabad multispectral image shape: ", featuresHyderabad.shape)

PratyushTripathy

from pyrsgis.convert import changeDimension

featuresBangalore = changeDimension(featuresBangalore)
labelBangalore = changeDimension (labelBangalore)
featuresHyderabad = changeDimension(featuresHyderabad)
nBands = featuresBangalore.shape[1]
labelBangalore = (labelBangalore == 1).astype(int)

print("Bangalore multispectral image shape: ", featuresBangalore.shape)
print("Bangalore binary built-up image shape: ", labelBangalore.shape)

PratyushTripathy

from sklearn.model_selection import train_test_split

xTrain, xTest, yTrain, yTest = train_test_split(featuresBangalore, labelBangalore, test_size=0.4, random_state=42)

print(xTrain.shape)
print(yTrain.shape)

print(xTest.shape)
print(yTest.shape)

PratyushTripathy

# Normalise the data
xTrain = xTrain / 255.0
xTest = xTest / 255.0
featuresHyderabad = featuresHyderabad / 255.0

# Reshape the data
xTrain = xTrain.reshape((xTrain.shape[0], 1, xTrain.shape[1]))
xTest = xTest.reshape((xTest.shape[0], 1, xTest.shape[1]))
featuresHyderabad = featuresHyderabad.reshape((featuresHyderabad.shape[0], 1, featuresHyderabad.shape[1]))

PratyushTripathy

from tensorflow import keras

# Define the parameters of the model
model = keras.Sequential([
    keras.layers.Flatten(input_shape=(1, nBands)),
    keras.layers.Dense(14, activation='relu'),
    keras.layers.Dense(2, activation='softmax')])

# Define the accuracy metrics and parameters
model.compile(optimizer="adam", loss="sparse_categorical_crossentropy", metrics=["accuracy"])

PratyushTripathy

from sklearn.metrics import confusion_matrix, precision_score, recall_score

# Predict for test data 
yTestPredicted = model.predict(xTest)
yTestPredicted = yTestPredicted[:,1]

# Calculate and display the error metrics
yTestPredicted = (yTestPredicted>0.5).astype(int)
cMatrix = confusion_matrix(yTest, yTestPredicted)
pScore = precision_score(yTest, yTestPredicted)

PratyushTripathy

predicted = model.predict(featuresHyderabad)
predicted = predicted[:,1]

#Export raster
prediction = np.reshape(predicted, (ds.RasterYSize, ds.RasterXSize))
outFile = 'Hyderabad_2011_BuiltupNN_predicted.tif'
raster.export(prediction, ds3, filename=outFile, dtype='float')

PratyushTripathy

import os
from pyrsgis import raster

# Change the directory
os.chdir("E:\\BuiltUpPrediction")

# Assign file names
mxBangalore = 'l5_Bangalore2011_raw.tif'
builtupBangalore = 'l5_Bangalore2011_builtup.tif'
mxHyderabad = 'l5_Hyderabad2011_raw.tif'

PratyushTripathy

import os, cv2
from pyrsgis import raster
import numpy as np

inFile = r"E:\CNN_Builtup\l5_Bangalore2011_raw.tif"
normFile = inFile.replace('.tif', '_normalised.tif')
histEqlFile = inFile.replace('.tif', '_histEqualised.tif')

ds, arr = raster.read(inFile)
normBands = np.random.randint(1, size = (ds.RasterCount, ds.RasterYSize, ds.RasterXSize))