added readme.md

index.py

@@ -2,33 +2,16 @@
 from rasterio.plot import show
 #import rasterio.plot as plt #import show
 import matplotlib.pyplot as plt
-from matplotlib import colors
 import numpy as np
 import pandas as pd
 plt.style.use('seaborn-deep')
 
 
-class MidpointNormalize(colors.Normalize):
-
-    def __init__(self, vmin=None, vmax=None, midpoint=None, clip=False):
-        self.midpoint = midpoint
-        colors.Normalize.__init__(self, vmin, vmax, clip)
-
-    def __call__(self, value, clip=None):
-
-        x, y = [self.vmin, self.midpoint, self.vmax], [0, 0.5, 1]
-        return np.ma.masked_array(np.interp(value, x, y), np.isnan(value))
-
-
 def index(b2, b1):
 
     logor = lambda x, y: np.logical_or(x>0, y>0)
     calc = lambda x, y, ch: np.where(ch, (x-y)/(x+y), 0)
-
-    #calculating NDVI
-    np.seterr(divide='ignore', invalid='ignore')
     ind = calc(b1, b2, logor(b2, b1))
-    print(ind.min(), ind.max())
     return ind
 
 
@@ -46,7 +29,7 @@ def plot_hist(plot_percent, dates, d):
         p = list(plot_percent[i])
         ax.bar(ind, p, bottom = bottom, width = 0.35, label = d[i])
         bottom = [bottom[i]+p[i] for i in range(len(p))]
-    ax.legend().set_draggable(True)
+
     ax.set_xticks(ind, minor=False)
     ax.set_xticklabels(dates)
     ax.set_title('Percent Area Cover for years {} \n'.format(', '.join(dates[:-1])+' and ' + dates[-1]))
@@ -103,7 +86,6 @@ def calc_table(ind, x, y, d):
 
     area.append(round(tot_area, 3))
     perc.append(100)
-    #df = pd.DataFrame({'Area(in m2)': area, 'Percentage Area': perc})
     return (perc, area)
 
 def diff_table(ind_arr, bands, dates, d):
@@ -151,7 +133,6 @@ def diff_table(ind_arr, bands, dates, d):
         date.append(str(last[i]))
         date.append('100%')
     df_dict['Total Area'] = date
-    print(df_dict)
     dfres = pd.DataFrame(df_dict)
 
     dfres.to_excel('result.xlsx')
@@ -168,7 +149,7 @@ def calc_diff(l, ch):
         bands.append([l[i][0].read(1), l[i][1].read(1),
                           l[i][2].read(1), l[i][3].read(1), trans[0], -trans[4]])
 
-    bands = list(map(lambda x: [i.astype(float) for i in x[:4]]+[x[4], x[5]], bands))
+
 
     if ch==1:
         ind_arr = [index(band[1], band[2]) for band in bands]
@@ -190,7 +171,7 @@ def calc_diff(l, ch):
 
     date_dict = plot_diff_util(ind_arr, bands, years, ch)
 
-    colormap = plt.cm.RdYlGn
+
 
     while 1:
 
@@ -199,9 +180,9 @@ def calc_diff(l, ch):
         mid = ind.mean()
         minx=np.nanmin(ind)
         maxx=np.nanmax(ind)
-        norm = MidpointNormalize(vmin=minx, vmax=maxx, midpoint=mid)
+
         ax = plt.subplot(121)
-        ax.imshow(ind, cmap=colormap, vmin=minx, vmax=maxx, norm=norm)
+        ax.imshow(ind, cmap=colormap, vmin=minx, vmax=maxx)
         ax.set_title('YEAR: {}'.format(date))
         ax1 = plt.subplot(122)
         ax1.imshow(band)
@@ -222,17 +203,4 @@ def calc_diff(l, ch):
     l = sorted(l, key = lambda x:x[4])
     calc_diff(l, ch)
 
-
-'''
-lsat/15/b3.tif lsat/15/b4.tif lsat/15/b5.tif lsat/15/b6.tif 2015
-lsat/19/b3.tif lsat/19/b4.tif lsat/19/b5.tif lsat/19/b6.tif 2019
-img1/8/b2.tif img1/8/b3.tif img1/8/b4.tif img1/8/b5.tif 2008          
-img1/11/b2.tif img1/11/b3.tif img1/11/b4.tif img1/11/b5.tif 2011
-img1/12/b2.tif img1/12/b3.tif img1/12/b4.tif img1/12/b5.tif 2012
-img1/13/b2.tif img1/13/b3.tif img1/13/b4.tif img1/13/b5.tif 2013
-img1/14/b2.tif img1/14/b3.tif img1/14/b4.tif img1/14/b5.tif 2014
-img1/15/b2.tif img1/15/b3.tif img1/15/b4.tif img1/15/b5.tif 2015
-img1/16/b2.tif img1/16/b3.tif img1/16/b4.tif img1/16/b5.tif 2016
-
-'''
 

lst.py

@@ -4,18 +4,6 @@
 import numpy as np
 from matplotlib import colors
 import seaborn as sns
-sns.set()
-
-class MidpointNormalize(colors.Normalize):
-
-    def __init__(self, vmin=None, vmax=None, midpoint=None, clip=False):
-        self.midpoint = midpoint
-        colors.Normalize.__init__(self, vmin, vmax, clip)
-
-    def __call__(self, value, clip=None):
-
-        x, y = [self.vmin, self.midpoint, self.vmax], [0, 0.5, 1]
-        return np.ma.masked_array(np.interp(value, x, y), np.isnan(value))
 
 
 normalize = lambda array: ((array - array.min())/(array.max() - array.min()))
@@ -35,26 +23,7 @@ def get_index(green, red, nir, swir):
 
 def get_LST(thermal, sat_name):
 
-    if sat_name == 7:
-        ml = 6.7087E-02
-        al = -0.06709
-        l_max = 17.040
-        l_min = 0.000
-        qcal_max = 255
-        qcal_min = 1
-        k1 = 666.09
-        k2 = 1282.71
-
-    else:
-        ml = 3.3420E-04
-        al = 0.10000
-        l_max = 22.00180
-        l_min = 0.10033
-        qcal_max = 65535
-        qcal_min = 1
-        k1 = 774.8853
-        k2 = 1321.0789
-
+    #k1, k2 qcal, ml, al values acquired from landsat data
     L_lambda = ml * thermal + al
     temp_cel = k2 / np.log(k1/L_lambda + 1) - 273.15
     return temp_cel
@@ -78,7 +47,7 @@ def show_corr(temp, l):
             print(e)
             pass
     temp = [i.mean() for i in temp]
-    print(temp, ndvi, ndwi, ndbi)
+
     ax = plt.subplot(131)
     ax.plot(temp, ndvi, marker = 'o')
     ax.set_title('LST vs NDVI')
@@ -116,11 +85,7 @@ def calc_everything(l):
     temp = calc_everything(l)
     show_corr(temp, l)
 
-'''
-lsat/00/b2.tif lsat/00/b3.tif lsat/00/b4.tif lsat/00/b5.tif lsat/00/b6.tif 7 2000
-lsat/15/b3.tif lsat/15/b4.tif lsat/15/b5.tif lsat/15/b6.tif lsat/15/b10.tif 8 2015
-lsat/19/b3.tif lsat/19/b4.tif lsat/19/b5.tif lsat/19/b6.tif lsat/19/b10.tif 8 2019
-'''
+
 
 
 

readme.md

@@ -0,0 +1,44 @@
+# README
+## About geoStudy
+
+- Remote sensing is the process of detecting and monitoring the physical traits of an area by measuring its reflected and emitted radiation at a distance from the targeted area. It is the acquisition of information about an object without making any physical contact with it. The data used in remote sensing can be categorized based on their resolution, electromagnetic spectrum, energy source, imaging media and number of bands. We decided to extract features like vegetation, water and built-up using the different indices like NDVI, NDWI and NDBI.
+
+- We decided to make a software that can be used calculate different features from satellite imagery. The software allows you to select raster imagery and find as well as compare the changes in different indices over a period of time. The software can also be used to study the effect of change in vegetation, water and built-up on Land Surface Temperature of the selected region.
+
+## An Insight
+
+- **Home Screen**
+
+![image](screenshots/1.png)
+
+
+- **NDVI Plot and False Color Composite of uploaded Raster Image**
+
+![image](screenshots/2.png)
+
+
+
+- **Area Cover Analysis of uploaded Raster Image**
+
+![image](screenshots/3.png)
+
+
+- **Percentage Area Cover for uploaded Images**
+
+![image](screenshots/5.png)
+
+
+
+- **Graph showing effect of change in all three indices on Land Surface Temperature**
+
+![image](screenshots/4.png)
+
+
+
+- The data was acquired from the [USGS portal](https://earthexplorer.usgs.gov) of NASA and [Bhuvan portal](https://bhuvan-app3.nrsc.gov.in/data/download/index.php) of ISRO
+
+- Find project video on: [YouTube](https://youtu.be/qfo3D78Pezc)
+
+
+
+