123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179 |
- #!/usr/bin/env python
- # class to convert/process modis data
- #
- # (c) Copyright Ingmar Nitze 2013
- # Authors: Ingmar Nitze, Luca Delucchi
- # Email: initze at ucc dot ie
- # Email: luca dot delucchi at iasma dot it
- #
- ##################################################################
- #
- # This MODIS Python class is licensed under the terms of GNU GPL 2.
- # This program is free software; you can redistribute it and/or
- # modify it under the terms of the GNU General Public License as
- # published by the Free Software Foundation; either version 2 of
- # the License, or (at your option) any later version.
- # This program is distributed in the hope that it will be useful,
- # but WITHOUT ANY WARRANTY; without even implied warranty of
- # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
- # See the GNU General Public License for more details.
- #
- ##################################################################
- """A class for the extraction and transformation of MODIS quality layers to
- specific information
- Classes:
- * :class:`QualityModis`
- """
- # python 2 and 3 compatibility
- from __future__ import print_function
- from builtins import dict
- import os
- try:
- import numpy as np
- except ImportError:
- raise ImportError('Numpy library not found, please install it')
- try:
- import osgeo.gdal as gdal
- import osgeo.gdal_array as gdal_array
- except ImportError:
- try:
- import gdal
- import gdal_array
- except ImportError:
- raise ImportError('Python GDAL library not found, please install '
- 'python-gdal')
- VALIDTYPES = dict({'13': list(map(str, list(range(1, 10)))), '11': list(map(str, list(range(1, 6))))})
- PRODUCTPROPS = dict({
- 'MOD13Q1': ([2], ['QAGrp1']),
- 'MYD13Q1': ([2], ['QAGrp1']),
- 'MOD13A1': ([2], ['QAGrp1']),
- 'MYD13A1': ([2], ['QAGrp1']),
- 'MOD13A2': ([2], ['QAGrp1']),
- 'MYD13A2': ([2], ['QAGrp1']),
- 'MOD13A3': ([2], ['QAGrp1']),
- 'MYD13A3': ([2], ['QAGrp1']),
- 'MOD13C1': ([2], ['QAGrp1']),
- 'MYD13C1': ([2], ['QAGrp1']),
- 'MOD13C2': ([2], ['QAGrp1']),
- 'MYD13C2': ([2], ['QAGrp1']),
- 'MOD11A1': ([1, 5], ['QAGrp2', 'QAGrp2']),
- 'MYD11A1': ([1, 5], ['QAGrp2', 'QAGrp2']),
- 'MOD11A2': ([1, 5], ['QAGrp4', 'QAGrp4']),
- 'MYD11A2': ([1, 5], ['QAGrp4', 'QAGrp4']),
- 'MOD11B1': ([1, 5, -2], ['QAGrp2', 'QAGrp2', 'QAGrp3']),
- 'MYD11B1': ([1, 5, -2], ['QAGrp2', 'QAGrp2', 'QAGrp3']),
- 'MOD11C1': ([1, 5, -2], ['QAGrp2', 'QAGrp2', 'QAGrp3']),
- 'MYD11C1': ([1, 5, -2], ['QAGrp2', 'QAGrp2', 'QAGrp3']),
- 'MOD11C2': ([1, 6], ['QAGrp2', 'QAGrp2']),
- 'MYD11C2': ([1, 6], ['QAGrp2', 'QAGrp2']),
- 'MOD11C3': ([1, 6], ['QAGrp2', 'QAGrp2']),
- 'MYD11C3': ([1, 6], ['QAGrp2', 'QAGrp2'])
- })
- QAindices = dict({
- 'QAGrp1': (16, [[-2, None], [-6, -2], [-8, -6], [-9, -8],
- [-10, -9], [-11, -10], [-14, -11], [-15, -14],
- [-16, -15]]),
- 'QAGrp2': (7, [[-2, None], [-3, -2], [-4, -3], [-6, -4],
- [-8, -6]]),
- 'QAGrp3': (7, [[-3, None], [-6, -3], [-7, -6]]),
- 'QAGrp4': (8, [[-2, None], [-4, -2], [-6, -4], [-8, -6]])
- })
- class QualityModis():
- """A Class for the extraction and transformation of MODIS
- quality layers to specific information
- :param str infile: the full path to the hdf file
- :param str outfile: the full path to the parameter file
- """
- def __init__(self, infile, outfile, qType=None, qLayer=None, pType=None):
- """Function to initialize the object"""
- self.infile = infile
- self.outfile = outfile
- self.qType = qType
- self.qLayer = qLayer
- self.qaGroup = None
- self.pType = pType
- def loadData(self):
- """loads the input file to the object"""
- os.path.isfile(self.infile)
- self.ds = gdal.Open(self.infile)
- def setProductType(self):
- """read productType from Metadata of hdf file"""
- if self.pType == None:
- self.productType = self.ds.GetMetadata()['SHORTNAME']
- else:
- self.productType = self.pType
- def setProductGroup(self):
- """read productGroup from Metadata of hdf file"""
- self.productGroup = self.productType[3:5]
- def setQAGroup(self):
- """set QA dataset group type"""
- if self.productType in list(PRODUCTPROPS.keys()):
- self.qaGroup = PRODUCTPROPS[self.productType][1][int(self.qLayer)-1]
- else:
- print("Product version is currently not supported!")
- def setQALayer(self):
- """function sets the input path of the designated QA layer"""
- self.qaLayer = self.ds.GetSubDatasets()[PRODUCTPROPS[self.productType][0][int(self.qLayer)-1]][0]
- def loadQAArray(self):
- """loads the QA layer to the object"""
- self.qaArray = gdal_array.LoadFile(self.qaLayer)
- def qualityConvert(self, modisQaValue):
- """converts encoded Bit-Field values to designated QA information"""
- startindex = QAindices[self.qaGroup][1][int(self.qType)-1][0]
- endindex = QAindices[self.qaGroup][1][int(self.qType)-1][1]
- return int(np.binary_repr(modisQaValue, QAindices[self.qaGroup][0])[startindex: endindex], 2)
- def exportData(self):
- """writes calculated QA values to physical .tif file"""
- qaDS = gdal.Open(self.qaLayer)
- dr = gdal.GetDriverByName('GTiff')
- outds = dr.Create(self.outfile, self.ncols, self.nrows, 1, gdal.GDT_Byte)
- outds.SetProjection(qaDS.GetProjection())
- outds.SetGeoTransform(qaDS.GetGeoTransform())
- outds.GetRasterBand(1).WriteArray(self.qaOut)
- outds = None
- qaDS = None
- def run(self):
- """Function defines the entire process"""
- self.loadData()
- self.setProductType()
- self.setProductGroup()
- #self.setDSversion()
- self.setQAGroup()
- self.setQALayer()
- self.loadQAArray()
- self.nrows, self.ncols = self.qaArray.shape
- print("Conversion started !")
- self.qaOut = np.zeros_like(self.qaArray, dtype=np.int8)
- if self.productGroup in ['11', '13'] and self.qType in VALIDTYPES[self.productGroup] and self.qaGroup != None:
- for val in np.unique(self.qaArray):
- ind = np.where(self.qaArray == val)
- self.qaOut[ind] = self.qualityConvert(self.qaArray[ind][0])
- self.exportData()
- print("Export finished!")
- else:
- print("This MODIS type is currently not supported.")
|