Landsat 8 Color Image Processing

This notebook is part of the collection accompanying the talk "Analyzing Satellite Images With Python Scientific Stack" by Milos Miljkovic given at PyData NYC 2014. The content is BSD licensed.

In [1]:
# To install watermark extension execute:
# %install_ext
%load_ext watermark
In [2]:
%watermark -v -m -p numpy,scikit-image,matplotlib
CPython 3.4.2
IPython 2.3.1

numpy 1.9.1
scikit-image 0.10.1
matplotlib 1.4.2

compiler   : GCC 4.2.1 (Apple Inc. build 5577)
system     : OS X 10.10.1
release    : 14.0.0
machine    : x86_64
processor  : Intel(R) Core(TM) i7-3740QM CPU @ 2.70GHz
CPU cores  : 8
interpreter: 64bit
In [3]:
import numpy as np
from skimage import io, exposure
from matplotlib import pyplot as plt
from IPython.display import Image
In [4]:
%matplotlib inline
In [5]:
def read_band(n):
    Load Landsat 8 band
    n - integer in the range 1-11
    img - 2D array of uint16 type
    if n in range(1, 12):
        tif_list = !ls *.TIF
        band_name = 'B' + str(n) + '.TIF'
        img_idx = [idx for idx, band_string in enumerate(tif_list) if band_name in band_string]
        img = io.imread(tif_list[img_idx[0]])
        return img
        print('Band number has to be in the range 1-11!')
In [6]:
def color_image_show(img, title):
    Show image
    img - 3D array of uint16 type
    title - string
    fig = plt.figure(figsize=(10, 10))
In [7]:
cd /Users/kronos/gis/l8/guinea_bissau/

Load RGB Bands

Landsat 8 bands corresponding to RGB are 4-3-2. Data is loaded as 2D uint16 arrays, then stacked into NumPy 3D array of the same type, and imaged.
Please note that matplotlib cannot display RGB uint16 arrays. Data is transformed to float64 in range 0-1.

In [8]:
b2 = read_band(2)
b3 = read_band(3)
b4 = read_band(4)

img432 = np.dstack((b4, b3, b2))
%reset_selective -f b
In [9]:
color_image_show(img432, '4-3-2 image, data set LC82040522013123LGN01')

Histogram Equalization

Histograms of RGB colors corresponding to raw data show that the data is not utilizing full 16-bit range (0-65535) afforded by the detector. Limits for all three colors are picked and data is rescaled. This results in apparent brightening of the image.

In [10]:
fig = plt.figure(figsize=(10, 7))

for color, channel in zip('rgb', np.rollaxis(img432, axis=-1)):
    counts, centers = exposure.histogram(channel)
    plt.plot(centers[1::], counts[1::], color=color)
In [11]:
img432_ha = np.empty(img432.shape, dtype='uint16')
lims = [(7100,14500), (8200, 14000), (9200,13500)]
for lim, channel in zip(lims, range(3)):
    img432_ha[:, :, channel] = exposure.rescale_intensity(img432[:, :, channel], lim)

color_image_show(img432_ha, '4-3-2 image, histogram equilized')