inex/isak
1
0
Fork 1
Code Issues Pull requests Projects Releases Wiki Activity

Add FAPAR

Browse source
This commit is contained in:
Inex Code 2021-12-17 21:30:42 +03:00
parent 3770c4dc2f
commit b92c7be4dc
2 changed files with 152 additions and 33 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

BIN
fapar.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 689 KiB

185
sat.py
View file

@ -1,7 +1,8 @@
from PIL import Image
import requests
from sat7_pointer import *
from numpy import np
import numpy as np
from tqdm import tqdm
# Load Landsat 7 band 1, 2, 3 TIF images and create a composite image
# from the three bands.
@ -14,17 +15,17 @@ from numpy import np
#
# The composite image is saved as a PNG file.
band1 = Image.open('LE07_L1TP_177025_20210723_20210818_02_T1_B1.TIF')
band2 = Image.open('LE07_L1TP_177025_20210723_20210818_02_T1_B2.TIF')
band3 = Image.open('LE07_L1TP_177025_20210723_20210818_02_T1_B3.TIF')
band1 = Image.open("LE07_L1TP_177025_20210723_20210818_02_T1_B1.TIF")
band2 = Image.open("LE07_L1TP_177025_20210723_20210818_02_T1_B2.TIF")
band3 = Image.open("LE07_L1TP_177025_20210723_20210818_02_T1_B3.TIF")
composite = Image.merge('RGB', (band3, band2, band1))
composite = Image.merge("RGB", (band3, band2, band1))
# Load corner coordinates of the image.
#
# The coordinates are stored in a MTL text file.
mtl_data = load_metadata('LE07_L1TP_177025_20210723_20210818_02_T1_MTL.txt')
mtl_data = load_metadata("LE07_L1TP_177025_20210723_20210818_02_T1_MTL.txt")
# Fetch coordinates of a city.
#
@ -35,14 +36,18 @@ mtl_data = load_metadata('LE07_L1TP_177025_20210723_20210818_02_T1_MTL.txt')
#
# # City is Belgorod, Russia.
url = 'http://nominatim.openstreetmap.org/search?q=Belgorod,+Russia&format=json'
url = "http://nominatim.openstreetmap.org/search?q=Belgorod,+Russia&format=json"
response = requests.get(url)
data = response.json()
# Convert bounding box coordinates to image coordinates.
x0, y1 = lat_lot_to_pixel(data[0]['boundingbox'][0], data[0]['boundingbox'][2], mtl_data)
x1, y0 = lat_lot_to_pixel(data[0]['boundingbox'][1], data[0]['boundingbox'][3], mtl_data)
x0, y1 = lat_lot_to_pixel(
data[0]["boundingbox"][0], data[0]["boundingbox"][2], mtl_data
)
x1, y0 = lat_lot_to_pixel(
data[0]["boundingbox"][1], data[0]["boundingbox"][3], mtl_data
)
print(x0, y0)
print(x1, y1)
@ -51,7 +56,7 @@ print(composite.size)
cropped = composite.crop((x0, y0, x1, y1))
cropped.save('cropped.png')
cropped.save("cropped.png")
###############################################################################
# LAB 2
@ -60,7 +65,7 @@ cropped.save('cropped.png')
# Load landsat band 4 TIF image.
# Band 4 is near infrared.
band4 = Image.open('LE07_L1TP_177025_20210723_20210818_02_T1_B4.TIF')
band4 = Image.open("LE07_L1TP_177025_20210723_20210818_02_T1_B4.TIF")
# Claculate the NDVI.
#
@ -116,7 +121,8 @@ def get_color(value):
else:
return (0, 0, 0)
for x in range(ndvi.size[0]):
for x in tqdm(range(ndvi.size[0]), desc="NDVI"):
for y in range(ndvi.size[1]):
r = red.getpixel((x, y))
nir = ndvi.getpixel((x, y))
@ -125,7 +131,7 @@ for x in range(ndvi.size[0]):
else:
result.putpixel((x, y), get_color((nir - r) / (nir + r)))
result.save('ndvi.png')
result.save("ndvi.png")
# Calculate FAPAR (Fraction of Absorbed Photosynthetically Active Radiation)
@ -133,28 +139,141 @@ result.save('ndvi.png')
#
# Bands 1, 3, 4 are used.
# Coefficients of polynominal g_0
# l0,1 l0,2 l0,3 l0,4 l0,5 l0,6
# 0.27505 0.35511 −0.004 −0.322 0.299 −0.0131
l_0 = np.array([0.27505, 0.35511, -0.004, -0.322, 0.299, -0.0131])
solar_zenith_angle = np.radians(float(mtl_data["SUN_ELEVATION"]))
sensor_zenith_angle = np.radians(0)
sun_sensor_relative_azimuth = np.radians(float(mtl_data["SUN_AZIMUTH"]))
# Coefficients of polynominal g_1
# l1,1 l1,2 l1,3 l1,4 l1,5 l1,6 l1,7 l1,8 l1,9 l1,10 l1,11
# −10.036 −0.019804 0.55438 0.14108 12.494 0 0 0 0 0 1.0
l_1 = np.array([-10.036, -0.019804, 0.55438, 0.14108, 12.494, 0, 0, 0, 0, 0, 1.0])
gain = [float(mtl_data["RADIANCE_MULT_BAND_" + str(i)]) for i in [1, 3, 4]]
offset = [float(mtl_data["RADIANCE_ADD_BAND_" + str(i)]) for i in [1, 3, 4]]
# Coefficients of polynominal g_2
# l2,1 l2,2 l2,3 l2,4 l2,5 l2,6 l2,7 l2,8 l2,9 l2,10
# 0.42720 0.069884 −0.33771 0.24690 −1.0821 −0.30401 −1.1024 −1.2596 −0.31949 −1.4864
l_2 = np.array([0.42720, 0.069884, -0.33771, 0.24690, -1.0821, -0.30401, -1.1024, -1.2596, -0.31949, -1.4864])
dsol = float(mtl_data["EARTH_SUN_DISTANCE"])
l = [l_0, l_1, l_2]
pic = [0.643, 0.80760, 0.89472]
k = [0.76611, 0.63931, 0.81037]
theta = [-0.10055, -0.06156, -0.03924]
# Function to calculate ratio of polynominals
def g_n(n, x, y):
if n == 0:
return l_0[0] * y - l_0[1] * x + l_0[2] / (l_0[3] - x)**2 + (l_0[4] - y)**2 + l_0[5]
if n == 1:
return l_1[0]
k = [0.63931,0.81037, 0.76611]
pic = [0.80760, 0.89472, 0.643]
theta = [-0.06156, -0.03924, -0.10055]
l[n][0] * (x + l[n][1])**2 + l[n][2] * (y + l[n][3])**2 + l[n][4] * x * y / l[n][5] * (x + l[n][6])**2 + l
E0 = [1969, 1551, 1044]
cosg = np.cos(solar_zenith_angle) * np.cos(sensor_zenith_angle) + np.sin(
solar_zenith_angle
) * np.sin(sensor_zenith_angle) * np.cos(sun_sensor_relative_azimuth)
G = (
np.tan(solar_zenith_angle) ** 2
+ np.tan(sensor_zenith_angle) ** 2
- 2
* np.tan(solar_zenith_angle)
* np.tan(sensor_zenith_angle)
* np.cos(sun_sensor_relative_azimuth)
) ** 0.5
polynoms = np.array(
[
[0.27505, 0.35511, -0.004, -0.322, 0.299, -0.0131, 0, 0, 0, 0, 0],
[-10.036, -0.019804, 0.55438, 0.14108, 12.494, 0, 0, 0, 0, 0, 1],
[
0.42720,
0.069884,
-0.33771,
0.24690,
-1.0821,
-0.30401,
-1.1024,
-1.2596,
-0.31949,
-1.4864,
0,
],
]
)
blue = band1.copy().crop((x0, y0, x1, y1))
red = band3.copy().crop((x0, y0, x1, y1))
nir = band4.copy().crop((x0, y0, x1, y1))
result = cropped.copy()
f1 = [
((np.cos(solar_zenith_angle) * np.cos(sensor_zenith_angle)) ** (k[i] - 1))
/ (np.cos(solar_zenith_angle) + np.cos(sensor_zenith_angle)) ** (1 - k[i])
for i in range(3)
]
f2 = [
(1 - theta[i] ** 2) / (1 + 2 * theta[i] * cosg + theta[i] ** 2) ** (3 / 2)
for i in range(3)
]
f3 = [1 + (1 - pic[i]) / (1 + G) for i in range(3)]
F = [f1[i] * f2[i] * f3[i] for i in range(3)]
def get_color_fapar(value, rho):
if (0 < rho[0] and rho[0] < 0.257752) \
and (0 < rho[1] and rho[1] < 0.48407) \
and (0 < rho[2] and rho[2] < 0.683928) \
and (rho[0] <= rho[2]) \
and (rho[2] >= 1.26826*rho[1]):
return get_color(value)
if (rho[0] <= 0) or (rho[1] <= 0) or (rho[2] <= 0):
return (0, 0, 0)
if (rho[0] >= 0.257752) or (rho[1] >= 0.48407) or (rho[2] >= 0.683928):
return (255, 255, 255)
if (0 < rho[0] and rho[0] < 0.257752) \
and (0 < rho[1] and rho[1] < 0.48407) \
and (0 < rho[2] and rho[2] < 0.683928) \
and (rho[0] >= rho[2]):
return (0, 0, 255)
if (0 < rho[0] and rho[0] < 0.257752) \
and (0 < rho[1] and rho[1] < 0.48407) \
and (0 < rho[2] and rho[2] < 0.683928) \
and (rho[0] <= rho[2]) \
and (1.25*rho[1] > rho[2]):
return (255, 150, 150)
if (rho[1] < 0) or (rho[2] < 0):
return (0, 0, 0)
if value < 0 or value > 1:
return (0, 0, 0)
return (int(180.0 * (1 - value)), 255, 255)
for x in tqdm(range(result.size[0]), desc="FAPAR"):
for y in range(result.size[1]):
bands = [blue.getpixel((x, y)), red.getpixel((x, y)), nir.getpixel((x, y))]
rho_i = [
(
(np.pi * (gain[i] * bands[i] + offset[i]) * dsol ** 2)
/ (E0[i] * np.cos(sensor_zenith_angle))
)
/ F[i]
for i in range(3)
]
g1 = (
(polynoms[1, 0] * (rho_i[0] + polynoms[1, 1]) ** 2)
+ (polynoms[1, 2] * (rho_i[1] + polynoms[1, 3]) ** 2)
+ polynoms[1, 4] * rho_i[0] * rho_i[1]
) / (
polynoms[1, 5] * (rho_i[0] + polynoms[1, 6]) ** 2
+ polynoms[1, 7] * (rho_i[1] + polynoms[1, 8]) ** 2
+ polynoms[1, 9] * rho_i[0] * rho_i[1]
+ polynoms[1, 10]
)
g2 = (
(polynoms[2, 0] * (rho_i[0] + polynoms[2, 1]) ** 2)
+ (polynoms[2, 2] * (rho_i[2] + polynoms[2, 3]) ** 2)
+ polynoms[2, 4] * rho_i[0] * rho_i[2]
) / (
polynoms[2, 5] * (rho_i[0] + polynoms[2, 6]) ** 2
+ polynoms[2, 7] * (rho_i[2] + polynoms[2, 8]) ** 2
+ polynoms[2, 9] * rho_i[0] * rho_i[2]
+ polynoms[2, 10]
)
FAPAR = ((polynoms[0, 0] * g2) - polynoms[0, 1] * g1 - polynoms[0, 2]) / (
(polynoms[0, 3] - g1) ** 2 + (polynoms[0, 4] - g2) ** 2 + polynoms[0, 5]
)
result.putpixel((x, y), get_color_fapar(FAPAR, rho_i))
result.save('fapar.png')