make_UTM_scene.py

# coding: utf-8

# 2017/04/13 Jin Katagi
import pandas as pd
import subprocess
import re


# In[2]:

def get_scene(id, category, lat, lon):
    """ return scene which include validation points.
    input:
        id(int)   : id of the validation points.
        lat(float): latitude of the validation points.
        lon(float): longitude of the validation point.
    return:
        tmp_df(data frame): df of the scene which include validation point.
    """
    
    # make empty datraframe.
    tmp_df = pd.DataFrame(columns=['id','category', 'lon', 'lat', 'scene'])
    
    # get scene name which include validatoin points.
    tmp_df["scene"] = df[(df["ll_lat"] < lat) & (lat < df["ul_lat"] ) &                  (df["ll_lon"] < lon) & (lon < df["ul_lon"])]["scene"]
    
    # append id, category, lat, lon
    tmp_df["id"] = id
    tmp_df["category"] =category 
    tmp_df["lat"] = lat 
    tmp_df["lon"] =lon 
    
    return tmp_df


# In[12]:

def run_gdal(file_path, file_name):
    gdal_command = ["gdalinfo", file_path + "/" + file_name]
    proc = subprocess.Popen(
    gdal_command,
    stdout=subprocess.PIPE)
    return proc

def run_grep(input_stdin):
    grep_command = ['grep', 'UTM']
    proc = subprocess.Popen(
        grep_command,
        stdin=input_stdin,
        stdout=subprocess.PIPE)
    return proc


# In[ ]:

def make_scene_zone(file_path,file_name, scene_zone_dict):
    """ get UTM zone and return pair dict of sceen and UTM zone number."""
    # run gdalinfo
    gdal_proc = run_gdal(file_path, file_name)
    # grep UTM
    grep_proc = run_grep(gdal_proc.stdout)

    out, err = grep_proc.communicate()
    string = out.decode('utf-8')

    pattern = r'.*(\d\d)N.*$'
    matchOB = re.match(pattern, string)
    
    if matchOB:
        scene_zone_dict[file_name] = matchOB.group(1)


# In[169]:

### Create dataframe df[id, category, lon, lat, scene] ###
# read validation csv failes
validation_df = pd.read_csv("./50points_coordinates_category.csv", header=None)

# read scene list.
AV2PAN_list_name  = '../AV2PAN/AV2PAN_list_decimal_sort.txt'
df = pd.read_csv(AV2PAN_list_name)

# create dataframe df['id, category, lon, lat, scene]
scene_df = pd.DataFrame(columns=['id','category', 'lon', 'lat', 'scene'])
for i, validation in validation_df.iterrows():
    # get scene name and append exit list
    scene_df = scene_df.append(get_scene(validation[0], validation[1], validation[3], validation[2]))
    
### create dict {'scene', 'UTM_zone} ###
scene_zone_dict = {}

file_path="/GEODATA/AV2PAN/PAN"

for file_name in scene_df['scene']:
    make_scene_zone(file_name, scene_zone_dict)
    
# convert to dataframe df[scene, UTM_zone] 
df = pd.DataFrame()
df['scene'] = scene_zone_dict.keys()
df['UTM_zone'] = scene_zone_dict.values()

# merge df[id, category, lon, lat, scene]  and df[scene, UTM_zone]
final_df = pd.merge(scene_df, df, how='inner')

#### save as csv. ###
final_df.to_csv(path_or_buf='./AV2PAN_UTM_zone.csv', sep=',', index=False )