slstr-image.py

import matplotlib.pyplot as plt
from pathlib import Path, PurePath
from image import ImageLoader, Regrid
from skimage.transform import resize
from skimage.exposure import exposure
import numpy as np
import sys
from PIL import Image as PillowImage
from PIL import ImageEnhance
import skimage

print('Path is ')
path = 'data/S3A_SL_1_RBT____20201120T100620_20201120T100920_20201121T162430_0179_065_179_1980_LN2_O_NT_004.SEN3/'
print('---')
product = ImageLoader(path)
# Get Channels
s1 = product.load_reflectance_channel(path, 1, 'an').to_array().values[0]
s3 = product.load_reflectance_channel(path, 3, 'an').to_array().values[0]
s5 = product.load_reflectance_channel(path, 5, 'an').to_array().values[0]

# Deal with NaNs
s1 = np.where(np.isnan(s1), 0, s1)
s3 = np.where(np.isnan(s3), 0, s3)
s5 = np.where(np.isnan(s5), 0, s5)

# Clip pixels > 1 to 1. This keeps the image in the range 0-1
s1 = np.where(s1 > 1., 1., s1)
s3 = np.where(s3 > 1., 1., s3)
s5 = np.where(s5 > 1., 1., s5)

# Construct Image
img = np.zeros((s1.shape[0], s1.shape[1], 3))
img[..., 0] = np.minimum(1, (s3 + s5) / 2)
img[..., 1] = np.minimum(1, s3)
img[..., 2] = np.minimum(1, (s3 + s1) / 2)

#Produce Image
img = skimage.exposure.rescale_intensity(img, out_range=(0, 255))

rgbImage = PillowImage.fromarray(np.uint8(img))
rgbImage.save('out.jpg')