ISCE_INSAR/components/isceobj/Sensor/ScanSAR/ALOS2.py

1028 lines
37 KiB
Python
Raw Normal View History

2019-01-16 19:40:08 +00:00
#!/usr/bin/env python3
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# Copyright 2010 California Institute of Technology. ALL RIGHTS RESERVED.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# United States Government Sponsorship acknowledged. This software is subject to
# U.S. export control laws and regulations and has been classified as 'EAR99 NLR'
# (No [Export] License Required except when exporting to an embargoed country,
# end user, or in support of a prohibited end use). By downloading this software,
# the user agrees to comply with all applicable U.S. export laws and regulations.
# The user has the responsibility to obtain export licenses, or other export
# authority as may be required before exporting this software to any 'EAR99'
# embargoed foreign country or citizen of those countries.
#
# Author: Walter Szeliga
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
import os
import datetime
import isceobj.Sensor.CEOS as CEOS
import logging
from isceobj.Orbit.Orbit import StateVector,Orbit
from isceobj.Planet.AstronomicalHandbook import Const
from isceobj.Planet.Planet import Planet
from iscesys.Component.Component import Component
from isceobj.Sensor import xmlPrefix
from isceobj.Util import Polynomial
from iscesys.DateTimeUtil import secondsSinceMidnight
import numpy as np
import struct
INPUT_DIRECTORY_LIST = Component.Parameter(
'inputDirList',
public_name='input directory',
type = str,
container = list,
mandatory = True,
doc = 'List of input directories to parse')
INPUT_SWATH_LIST = Component.Parameter(
'swaths',
public_name='swaths',
type=int,
container=list,
mandatory=False,
default=None,
doc = 'List of swaths to use')
POLARIZATION = Component.Parameter(
'polarization',
public_name='polarization',
type=str,
default='hh',
mandatory=False,
doc='Polarization to search for')
VIRTUAL_FILES = Component.Parameter(
'virtualFiles',
public_name='use virtual files',
type=bool,
default=True,
mandatory=False,
doc='Use virtual files instead of using disk space')
OUTPUTDIR = Component.Parameter(
'output',
public_name='output directory',
type = str,
default=None,
mandatory = True,
doc = 'Output directory for unpacking the data')
MAX_SWATHS = Component.Parameter(
'maxSwaths',
public_name='maximum number of swaths',
type=int,
default=5,
mandatory=True,
doc = 'Maximum number of swaths to scan for')
####ALOS2 directory browser
class ALOS2Scanner(Component):
family = 'alos2scanner'
parameter_list = (INPUT_DIRECTORY_LIST,
INPUT_SWATH_LIST,
POLARIZATION,
VIRTUAL_FILES,
OUTPUTDIR,
MAX_SWATHS)
modes = ['WBS', 'WBD', 'WWS', 'WWD', 'VBS', 'VBD']
def __init__(self, name=''):
super(ALOS2Scanner, self).__init__(family=self.__class__.family, name=name)
def scan(self):
if isinstance(self.inputDirList, str):
self.inputDirList = [self.inputDirList]
frames = []
for indir in self.inputDirList:
frames.append( self.scanDir(indir))
if len(frames) == 0:
raise Exception('No products found in the input directories')
###Estimate common swaths
return frames
def extractImage(self):
'''
Actual extraction of SLCs.
'''
totalSwaths = []
frames = self.scan()
###Currently assuming one frame
###Modify here for multiple frames
for swathid, img in frames[0].items():
sensor = ALOS2()
sensor.configure()
sensor._leaderFile = img['leaderfile']
sensor._imageFile = img['imgfile']
outdir = os.path.join(self.output, img['frame'], 's{0}'.format(swathid))
sensor.output = os.path.join(outdir, 'swath.slc')
if not os.path.isdir(outdir):
os.makedirs(outdir)
sensor.extractImage(virtual=self.virtualFiles)
sensor.extractDoppler()
sensor.refineBurstTiming()
totalSwaths.append(sensor.frame)
return totalSwaths
def scanDir(self, indir):
'''
Scan directory for IMG files.
'''
import glob
import os
imgFiles = glob.glob(os.path.join(indir, 'IMG-{0}-ALOS2*-*-*'.format(self.polarization.upper())))
###No IMG files found
if len(imgFiles) == 0:
return None
###Sort the filenames
imgFiles = sorted(imgFiles)
#######
wbFiles = []
for infile in imgFiles:
basefile = os.path.basename(infile)
##Check for each mode
for mode in self.modes:
if mode in basefile:
wbFiles.append(infile)
break
if len(wbFiles) == 0:
return None
###Check if user has requested specific files
frames = []
datatakes = []
imgmodes = []
for infile in wbFiles:
basefile = os.path.basename(infile)
frames.append( basefile.split('-')[2][-4:])
datatakes.append( basefile.split('-')[2][5:10])
imgmodes.append( basefile.split('-')[-2][0:3])
if any([x!=frames[0] for x in frames]):
print('Multiple frames found in same dir')
print(set(frames))
raise Exception('Multiple ALOS2 frames in same dir')
if any([x!=datatakes[0] for x in datatakes]):
print('Multiple datatakes found in same dir')
print(set(datatakes))
raise Exception('Multiple ALOS2 datatakes found in same dir')
if any([x!=imgmodes[0] for x in imgmodes]):
print('Multiple imaging modes found in same dir')
print(set(imgmodes))
raise Exception('Multiple ALOS2 imaging modes found in same dir')
swaths = {}
for infile in wbFiles:
params = {}
params['datatake'] = datatakes[0]
params['frame'] = frames[0]
params['imgfile'] = infile
swathid = int(os.path.basename(infile)[-1])
##If user has requested specific swaths
if self.swaths:
if swathid in self.swaths:
swaths[swathid] = params
else:
swaths[swathid] = params
###Ensure that a LED file exists that matches the data
ldrfiles = glob.glob(os.path.join(indir, 'LED-ALOS2{0}{1}-*'.format(datatakes[0], frames[0])))
if len(ldrfiles) == 0:
raise Exception('No leader file found in ALOS2 directory')
if len(ldrfiles) > 1:
raise Exception('More than one leader file found in ALOS2 directory')
leaderFile = ldrfiles[0]
for key, val in swaths.items():
swaths[key]['leaderfile'] = leaderFile
return swaths
#####Actual ALOS reader
#Sometimes the wavelength in the meta data is not correct.
#If the user sets this parameter, then the value in the
#meta data file is ignored.
WAVELENGTH = Component.Parameter(
'wavelength',
public_name='radar wavelength',
default=None,
type=float,
mandatory=False,
doc='Radar wavelength in meters.'
)
LEADERFILE = Component.Parameter(
'_leaderFile',
public_name='leaderfile',
default=None,
type=str,
mandatory=True,
doc='Name of the leaderfile.'
)
IMAGEFILE = Component.Parameter(
'_imageFile',
public_name='imagefile',
default=None,
type=str,
mandatory=True,
doc='Name of the imagefile.'
)
OUTPUT = Component.Parameter('output',
public_name='OUTPUT',
default=None,
type=str,
doc = 'Directory where bursts get unpacked')
###List of facilities
FRAME = Component.Facility('frame',
public_name = 'frame',
module = 'isceobj.Sensor.ScanSAR',
factory = 'createFullApertureSwathSLCProduct',
args = (),
mandatory=True,
doc = 'Full aperture swath slc product populated by the reader')
class ALOS2(Component):
"""
Code to read CEOSFormat leader files for ALOS2 SLC data.
"""
family = 'alos2'
parameter_list = (WAVELENGTH,
LEADERFILE,
IMAGEFILE,
OUTPUT)
facility_list = (FRAME,)
fsampConst = { 104: 1.047915957140240E+08,
52: 5.239579785701190E+07,
34: 3.493053190467460E+07,
17: 1.746526595233730E+07 }
#Orbital Elements (Quality) Designator
#ALOS-2/PALSAR-2 Level 1.1/1.5/2.1/3.1 CEOS SAR Product Format Description
#PALSAR-2_xx_Format_CEOS_E_r.pdf
orbitElementsDesignator = {'0':'preliminary',
'1':'decision',
'2':'high precision'}
def __init__(self, name=''):
super().__init__(family=self.__class__.family, name=name)
self.leaderFile = None
self.imageFile = None
#####Soecific doppler functions for ALOS2
self.doppler_coeff = None
self.azfmrate_coeff = None
self.lineDirection = None
self.pixelDirection = None
self.constants = {'polarization': 'HH',
'antennaLength': 10}
def getFrame(self):
return self.frame
def parse(self):
self.leaderFile = LeaderFile(self, file=self._leaderFile)
self.leaderFile.parse()
self.imageFile = ImageFile(self, file=self._imageFile)
self.imageFile.parse()
self.populateMetadata()
def populateMetadata(self):
"""
Create the appropriate metadata objects from our CEOSFormat metadata
"""
frame = self._decodeSceneReferenceNumber(self.leaderFile.sceneHeaderRecord.metadata['Scene reference number'])
fsamplookup = int(self.leaderFile.sceneHeaderRecord.metadata['Range sampling rate in MHz'])
rangePixelSize = Const.c/(2*self.fsampConst[fsamplookup])
ins = self.frame.getInstrument()
platform = ins.getPlatform()
platform.setMission(self.leaderFile.sceneHeaderRecord.metadata['Sensor platform mission identifier'])
platform.setAntennaLength(self.constants['antennaLength'])
platform.setPointingDirection(1)
platform.setPlanet(Planet(pname='Earth'))
if self.wavelength:
ins.setRadarWavelength(float(self.wavelength))
# print('ins.radarWavelength = ', ins.getRadarWavelength(),
# type(ins.getRadarWavelength()))
else:
ins.setRadarWavelength(self.leaderFile.sceneHeaderRecord.metadata['Radar wavelength'])
ins.setIncidenceAngle(self.leaderFile.sceneHeaderRecord.metadata['Incidence angle at scene centre'])
self.frame.getInstrument().setPulseRepetitionFrequency(self.leaderFile.sceneHeaderRecord.metadata['Pulse Repetition Frequency in mHz']*1.0e-3)
ins.setRangePixelSize(rangePixelSize)
ins.setRangeSamplingRate(self.fsampConst[fsamplookup])
ins.setPulseLength(self.leaderFile.sceneHeaderRecord.metadata['Range pulse length in microsec']*1.0e-6)
chirpSlope = self.leaderFile.sceneHeaderRecord.metadata['Nominal range pulse (chirp) amplitude coefficient linear term']
chirpPulseBandwidth = abs(chirpSlope * self.leaderFile.sceneHeaderRecord.metadata['Range pulse length in microsec']*1.0e-6)
ins.setChirpSlope(chirpSlope)
ins.setInPhaseValue(7.5)
ins.setQuadratureValue(7.5)
self.lineDirection = self.leaderFile.sceneHeaderRecord.metadata['Time direction indicator along line direction'].strip()
self.pixelDirection = self.leaderFile.sceneHeaderRecord.metadata['Time direction indicator along pixel direction'].strip()
######ALOS2 includes this information in clock angle
clockAngle = self.leaderFile.sceneHeaderRecord.metadata['Sensor clock angle']
if clockAngle == 90.0:
platform.setPointingDirection(-1)
elif clockAngle == -90.0:
platform.setPointingDirection(1)
else:
raise Exception('Unknown look side. Clock Angle = {0}'.format(clockAngle))
# print(self.leaderFile.sceneHeaderRecord.metadata["Sensor ID and mode of operation for this channel"])
self.frame.setFrameNumber(frame)
self.frame.setOrbitNumber(self.leaderFile.sceneHeaderRecord.metadata['Orbit number'])
self.frame.setProcessingFacility(self.leaderFile.sceneHeaderRecord.metadata['Processing facility identifier'])
self.frame.setProcessingSystem(self.leaderFile.sceneHeaderRecord.metadata['Processing system identifier'])
self.frame.setProcessingSoftwareVersion(self.leaderFile.sceneHeaderRecord.metadata['Processing version identifier'])
self.frame.setPolarization(self.constants['polarization'])
self.frame.setNumberOfLines(self.imageFile.imageFDR.metadata['Number of lines per data set'])
self.frame.setNumberOfSamples(self.imageFile.imageFDR.metadata['Number of pixels per line per SAR channel'])
######
orb = self.frame.getOrbit()
orb.setOrbitSource('Header')
orb.setOrbitQuality(
self.orbitElementsDesignator[
self.leaderFile.platformPositionRecord.metadata['Orbital elements designator']
]
)
t0 = datetime.datetime(year=self.leaderFile.platformPositionRecord.metadata['Year of data point'],
month=self.leaderFile.platformPositionRecord.metadata['Month of data point'],
day=self.leaderFile.platformPositionRecord.metadata['Day of data point'])
t0 = t0 + datetime.timedelta(seconds=self.leaderFile.platformPositionRecord.metadata['Seconds of day'])
#####Read in orbit in inertial coordinates
deltaT = self.leaderFile.platformPositionRecord.metadata['Time interval between data points']
numPts = self.leaderFile.platformPositionRecord.metadata['Number of data points']
orb = self.frame.getOrbit()
for i in range(numPts):
vec = StateVector()
t = t0 + datetime.timedelta(seconds=i*deltaT)
vec.setTime(t)
dataPoints = self.leaderFile.platformPositionRecord.metadata['Positional Data Points'][i]
pos = [dataPoints['Position vector X'], dataPoints['Position vector Y'], dataPoints['Position vector Z']]
vel = [dataPoints['Velocity vector X'], dataPoints['Velocity vector Y'], dataPoints['Velocity vector Z']]
vec.setPosition(pos)
vec.setVelocity(vel)
orb.addStateVector(vec)
###This is usually available with ALOS SLC data.
###Unfortunately set to all zeros for ScanSAR data
#self.doppler_coeff = [self.leaderFile.sceneHeaderRecord.metadata['Cross track Doppler frequency centroid constant term'],
#self.leaderFile.sceneHeaderRecord.metadata['Cross track Doppler frequency centroid linear term'],
#self.leaderFile.sceneHeaderRecord.metadata['Cross track Doppler frequency centroid quadratic term']]
self.azfmrate_coeff = [self.leaderFile.sceneHeaderRecord.metadata['Cross track Doppler frequency rate constant term'],
self.leaderFile.sceneHeaderRecord.metadata['Cross track Doppler frequency rate linear term'],
self.leaderFile.sceneHeaderRecord.metadata['Cross track Doppler frequency rate quadratic term']]
###Reading in approximate values instead
###Note that these are coeffs vs slant range in km
self.doppler_coeff = [self.leaderFile.sceneHeaderRecord.metadata['Doppler center frequency constant term'],
self.leaderFile.sceneHeaderRecord.metadata['Doppler center frequency linear term']]
# print('Terrain height: ', self.leaderFile.sceneHeaderRecord.metadata['Average terrain ellipsoid height'])
def extractImage(self, virtual=False):
import isceobj
if (self.imageFile is None) or (self.leaderFile is None):
self.parse()
###Generating XML file first as renderHdr also creates a VRT
###We want the virtual CEOS VRT to overwrite the general style VRT
rawImage = isceobj.createSlcImage()
rawImage.setByteOrder('l')
rawImage.setFilename(self.output)
rawImage.setAccessMode('read')
rawImage.setWidth(self.imageFile.width)
rawImage.setLength(self.imageFile.length)
rawImage.setXmin(0)
rawImage.setXmax(self.imageFile.width)
rawImage.renderHdr()
self.frame.setImage(rawImage)
self.imageFile.extractImage(output=self.output, virtual=virtual)
self.frame.setSensingStart(self.imageFile.sensingStart)
self.frame.setSensingStop(self.imageFile.sensingStop)
sensingMid = self.imageFile.sensingStart + datetime.timedelta(seconds = 0.5* (self.imageFile.sensingStop - self.imageFile.sensingStart).total_seconds())
self.frame.setSensingMid(sensingMid)
self.frame.setStartingRange(self.imageFile.nearRange)
self.frame.getInstrument().setPulseRepetitionFrequency(self.imageFile.prf)
pixelSize = self.frame.getInstrument().getRangePixelSize()
farRange = self.imageFile.nearRange + (pixelSize-1) * self.imageFile.width
self.frame.setFarRange(farRange)
return
def extractDoppler(self):
'''
Evaluate the doppler and fmrate polynomials.
'''
import copy
##We typically use this ALOS2 SLCs
####CEOS already provides function vs pixel
#self.frame._dopplerVsPixel = self.doppler_coeff
##Instead for ScanSAR data, we have to do the mapping from approx coeffs
frame = self.frame
width = frame.getNumberOfSamples()
rng = frame.startingRange + np.arange(0,width,100) * 0.5 * Const.c/frame.rangeSamplingRate
doppler = self.doppler_coeff[0] + self.doppler_coeff[1] * rng/1000.
dfit = np.polyfit( np.arange(0, width, 100), doppler, 1)
self.frame._dopplerVsPixel=[dfit[1], dfit[0], 0., 0.]
##We have to compute FM rate here.
##Cunren's observation that this is all set to zero in CEOS file.
##Simplification from Cunren's fmrate.py script
##Should be the same as the one in focus.py
planet = self.frame.instrument.platform.planet
elp = copy.copy(planet.ellipsoid)
svmid = self.frame.orbit.interpolateOrbit(self.frame.sensingMid, method='hermite')
xyz = svmid.getPosition()
vxyz = svmid.getVelocity()
llh = elp.xyz_to_llh(xyz)
hdg = self.frame.orbit.getENUHeading(self.frame.sensingMid)
elp.setSCH(llh[0], llh[1], hdg)
sch, schvel = elp.xyzdot_to_schdot(xyz, vxyz)
##Computeation of acceleration
dist= np.linalg.norm(xyz)
r_spinvec = np.array([0., 0., planet.spin])
r_tempv = np.cross(r_spinvec, xyz)
inert_acc = np.array([-planet.GM*x/(dist**3) for x in xyz])
r_tempa = np.cross(r_spinvec, vxyz)
r_tempvec = np.cross(r_spinvec, r_tempv)
axyz = inert_acc - 2 * r_tempa - r_tempvec
schbasis = elp.schbasis(sch)
schacc = np.dot(schbasis.xyz_to_sch, axyz).tolist()[0]
##Jumping back straight into Cunren's script here
centerVel = schvel
centerAcc = schacc
avghgt = llh[2]
radiusOfCurvature = elp.pegRadCur
frame = self.frame
fmrate = []
width = self.frame.getNumberOfSamples()
lookSide = self.frame.instrument.platform.pointingDirection
centerVelNorm = np.linalg.norm(centerVel)
##Retaining Cunren's code for computing at every pixel.
##Can be done every 10th pixel since we only fit a quadratic/ cubic.
##Also can be vectorized for speed.
for ii in range(width):
rg = frame.startingRange + ii * 0.5 * Const.c / frame.rangeSamplingRate
dop = np.polyval(frame._dopplerVsPixel[::-1], ii)
th = np.arccos(((avghgt+radiusOfCurvature)**2 + rg**2 -radiusOfCurvature**2)/(2.0 * (avghgt + radiusOfCurvature) * rg))
thaz = np.arcsin(((frame.radarWavelegth*dop/(2.0*np.sin(th))) + (centerVel[2] / np.tan(th))) / np.sqrt(centerVel[0]**2 + centerVel[1]**2)) - lookSide * np.arctan(centerVel[1]/centerVel[0])
lookVec = [ np.sin(th) * np.sin(thaz),
np.sin(th) * np.cos(thaz) * lookSide,
-np.cos(th)]
vdotl = np.dot(lookVec, centerVel)
adotl = np.dot(lookVec, centerAcc)
fmratex = 2.0*(adotl + (vdotl**2 - centerVelNorm**2)/rg)/(frame.radarWavelegth)
fmrate.append(fmratex)
##Fitting order 2 polynomial to FM rate
p = np.polyfit(np.arange(width), fmrate,2)
frame._fmrateVsPixel = list(p[::-1])
def _decodeSceneReferenceNumber(self,referenceNumber):
return referenceNumber
def refineBurstTiming(self):
'''
This is combination of burst_time2.py and burst_time.py from Cunren.
'''
slc = self.frame.image.filename
##First pass of burst_time.py
delta_line = 15000
bursts1 = self.burst_time( slc,
firstLine=delta_line, firstPixel=1000)
##Number of burst cycles
num_nc = np.around((self.frame.getNumberOfLines() - delta_line*2)/ self.frame.ncraw)
###Second pass
start_line2 = np.around( delta_line + num_nc * self.frame.ncraw)
bursts2 = self.burst_time( slc,
firstLine=start_line2, firstPixel=1000)
###Check if there were differences
LineDiffIndex = 0
LineDiffMin = np.fabs( bursts1['estimatedStartLine'] + self.frame.ncraw * LineDiffIndex - bursts2['estimatedStartLine'])
for ii in range(100000):
LineDiffMinx = np.fabs(bursts1['estimatedStartLine'] + self.frame.ncraw * ii - bursts2['estimatedStartLine'])
if LineDiffMinx <= LineDiffMin:
LineDiffMin = LineDiffMinx
LineDiffIndex = ii
###Update correct burst cycle value
print('Burst cycle length before correction: ', self.frame.ncraw)
self.frame.ncraw = self.frame.ncraw - (bursts1['estimatedStartLine'] + self.frame.ncraw * LineDiffIndex - bursts2['estimatedStartLine'])/LineDiffIndex
print('Burst cycle length after correction: ', self.frame.ncraw)
###Final run with updated burst cycle length
start_line1 = np.around(self.frame.getNumberOfLines() / 2.0)
bursts = self.burst_time( slc,
firstLine=start_line1, firstPixel=1000)
self.frame.burstStartLines = bursts['startLines']
for ii, val in enumerate(self.frame.burstStartLines):
print('Burst: {0}, Line: {1}'.format(ii, val))
def burst_time(self, slcfile,
firstLine=500, firstPixel=500,
nRange=400):
'''
Generates a linear FM signal and returns correlation with signal.
'''
def create_lfm(ns, it, offset, k):
'''
Create linear FM signal.
ns: Number of samples
it: Time interval of samples
offset: offset
k: linear FM rate
'''
ht = (ns-1)/2.0
t = np.arange(-ht, ht+1.0, 1)
t = (t + offset) * it
lfm = np.exp(1j * np.pi * k * t**2)
return lfm
from osgeo import gdal
frame = self.frame
width = frame.getNumberOfSamples()
length = frame.getNumberOfLines()
prf = frame.PRF
nb = frame.nbraw
nc = frame.ncraw
fmrateCoeff = frame._fmrateVsPixel
sensing_start = frame.getSensingStart()
###Using convention that Fmrate is positive
ka = -np.polyval(fmrateCoeff[::-1], np.arange(width))
###Area to be used for estimation
saz = firstLine #Startline to be included
naz = int(np.round(nc)) #Number of lines to be used
eaz = saz + naz-1 #Ending line to be used
caz = int(np.round((saz+eaz)/2.0)) #Central line of lines used
caz_deramp = (saz+eaz)/2.0 #Center of deramp signal
srg = firstPixel #Start column to be used
nrg = nRange #Number columns to be used
erg = srg + nrg - 1 #Ending column to be used
crg = int(np.round((srg+erg)/2.0)) #Central column
if not (saz >=0 and saz <= length-1):
raise Exception('Invalid starting line \n')
if not (eaz >=0 and eaz <= length-1):
raise Exception('Invalid ending line \n')
if not (srg >= 0 and erg <= width-1):
raise Exception('Invalid starting column \n')
if not (erg >=0 and erg <= width-1):
raise Exception('Invalid ending column \n')
###Calculate full aperture length
nFullAperture = int(np.round(prf/ka[crg]/(1.0/prf)))
nazfft = int(2**(int(np.ceil(np.log2(nFullAperture)))))
###Create the deramp function using fmrate
deramp = np.zeros((naz,nrg), dtype=np.complex64)
for ii in range(nrg):
deramp[:,ii] = create_lfm(naz, 1.0/prf, 0, -ka[ii+srg])
###Read in chunk of data
ds = gdal.Open(slcfile + '.vrt', gdal.GA_ReadOnly)
data = ds.ReadAsArray(srg, saz, nrg, naz)
ds = None
###deramp the data
datadr = deramp * data
#Compute spectrum
spec = np.fft.fft(datadr, n=nazfft, axis=0)
#Center the spectrum
spec = np.fft.fftshift(spec, axes=0)
##Average the spectrum
avgSpec = np.mean( np.abs(spec), axis=1)
###Number of bursts in freq domain
nbs = int(np.round(nb*(1.0/prf)*ka[crg]/prf*nazfft))
###Number of samples of the burst cycle in frequency domain
ncs = int(np.round(nc*(1.0/prf)*ka[crg]/prf*nazfft))
###Create a signal corresponding to 1 burst spectrum length
rect = np.ones(nbs, dtype=np.float32)
##Correlated burst with average spectrum
corr = np.correlate(avgSpec, rect, 'same')
###Find burst spectrum center
ncs_rh = int(np.round((nazfft - ncs)/2.0))
##Offset between spectrum center and center
offset_spec = np.argmax(corr[ncs_rh:ncs_rh+ncs]) + ncs_rh - (nazfft-1.0)/2.0
##Offset in azimuth lines
offset_naz = offset_spec / nazfft * prf / ka[crg] / (1.0 / prf)
##Starting line of the burst (fractional line number)
saz_burst = -offset_naz + caz_deramp - (nb-1.0)/2.0
####Find the start lines of all bursts
burstStartLines = []
burstStartTimes = []
for ii in range(-100000, 100000):
saz_burstx = saz_burst + nc * ii
if (saz_burstx >= 0.0) and (saz_burstx <= length):
st_burstx = sensing_start + datetime.timedelta(seconds=saz_burstx/prf)
burstStartLines.append(saz_burstx)
burstStartTimes.append(st_burstx)
bursts = {}
bursts['startLines'] = burstStartLines
bursts['startTimes'] = burstStartTimes
bursts['estimatedStartLine'] = saz_burst
#for ii in range(len(bursts['startLines'])):
# print(ii, bursts['startLines'][ii], bursts['startTimes'][ii])
return bursts
class LeaderFile(object):
def __init__(self, parent, file=None):
self.parent = parent
self.file = file
self.leaderFDR = None
self.sceneHeaderRecord = None
self.platformPositionRecord = None
def parse(self):
"""
Parse the leader file to create a header object
"""
try:
fp = open(self.file,'rb')
except IOError as errs:
errno,strerr = errs
print("IOError: %s" % strerr)
return
# Leader record
self.leaderFDR = CEOS.CEOSDB(xml=os.path.join(xmlPrefix,'alos2_slc/leader_file.xml'),dataFile=fp)
self.leaderFDR.parse()
fp.seek(self.leaderFDR.getEndOfRecordPosition())
# Scene Header
self.sceneHeaderRecord = CEOS.CEOSDB(xml=os.path.join(xmlPrefix,'alos2_slc/scene_record.xml'),dataFile=fp)
self.sceneHeaderRecord.parse()
fp.seek(self.sceneHeaderRecord.getEndOfRecordPosition())
# Platform Position
self.platformPositionRecord = CEOS.CEOSDB(xml=os.path.join(xmlPrefix,'alos2_slc/platform_position_record.xml'),dataFile=fp)
self.platformPositionRecord.parse()
fp.seek(self.platformPositionRecord.getEndOfRecordPosition())
#####Skip attitude information
fp.seek(16384,1)
#####Skip radiometric information
fp.seek(9860,1)
####Skip the data quality information
fp.seek(1620,1)
####Skip facility 1-4
fp.seek(325000 + 511000 + 3072 + 728000, 1)
####Read facility 5
self.facilityRecord = CEOS.CEOSDB(xml=os.path.join(xmlPrefix,'alos2_slc/facility_record.xml'), dataFile=fp)
self.facilityRecord.parse()
fp.close()
class VolumeDirectoryFile(object):
def __init__(self,file=None):
self.file = file
self.metadata = {}
def parse(self):
try:
fp = open(self.file,'rb')
except IOError as errs:
errno,strerr = errs
print("IOError: %s" % strerr)
return
volumeFDR = CEOS.CEOSDB(xml=os.path.join(xmlPrefix,'alos2_slc/volume_descriptor.xml'),dataFile=fp)
volumeFDR.parse()
fp.seek(volumeFDR.getEndOfRecordPosition())
fp.close()
class ImageFile(object):
def __init__(self, parent, file=None):
self.parent = parent
self.file = file
self.imageFDR = None
self.sensingStart = None
self.sensingStop = None
self.nearRange = None
self.prf = None
self.image_record = os.path.join(xmlPrefix,'alos2_slc/image_record.xml')
self.logger = logging.getLogger('isce.sensor.alos2')
def parse(self):
try:
fp = open(self.file,'rb')
except IOError as errs:
errno,strerr = errs
print("IOError: %s" % strerr)
return
self.imageFDR = CEOS.CEOSDB(xml=os.path.join(xmlPrefix,'alos2_slc/image_file.xml'), dataFile=fp)
self.imageFDR.parse()
fp.seek(self.imageFDR.getEndOfRecordPosition())
self._calculateRawDimensions(fp)
fp.close()
def writeRawData(self, fp, line):
'''
Convert complex integer to complex64 format.
'''
cJ = np.complex64(1j)
data = line[0::2] + cJ * line[1::2]
data.tofile(fp)
def extractImage(self, output=None, virtual=False):
"""
Extract I and Q channels from the image file
"""
if virtual:
output = output + '.vrt'
else:
try:
output = open(output, 'wb')
except IOError as strerr:
raise Exceptin("IOError: {0}".format(strerr))
if self.imageFDR is None:
self.parse()
###Open the image file for reading
try:
fp = open(self.file, 'rb')
except IOError as strerr:
self.logger.error(" IOError: %s" % strerr)
return
fp.seek(self.imageFDR.getEndOfRecordPosition(),os.SEEK_SET)
offsetAfterImageFDR = fp.tell()
dataLen = self.imageFDR.metadata['Number of pixels per line per SAR channel']
self.width = dataLen
##Leaderfile PRF
prf = self.parent.leaderFile.sceneHeaderRecord.metadata['Pulse Repetition Frequency in mHz']*1.0e-3
#choose PRF according to operation mode. Cunren Liang, 2015
operationMode = "{}".format(self.parent.leaderFile.sceneHeaderRecord.metadata['Sensor ID and mode of operation for this channel'])
operationMode =operationMode[10:12]
if operationMode not in ['08', '09']:
# Operation mode
# '00': Spotlight mode
# '01': Ultra-fine
# '02': High-sensitive
# '03': Fine
# '08': ScanSAR nominal mode
# '09': ScanSAR wide mode
# '18': Full (Quad.) pol./High-sensitive
# '19': Full (Quad.) pol./Fine
print('This reader only supports ScanSAR full aperture data parsing.')
raise Exception('Use stripmap reader for other modes')
if operationMode != '08':
raise Exception('Only ScanSAR nominal mode is currently supported')
# Extract the I and Q channels
imageData = CEOS.CEOSDB(xml=self.image_record,dataFile=fp)
###If only a VRT needs to be written
for line in range(self.length):
if ((line%1000) == 0):
self.logger.debug("Extracting line %s" % line)
imageData.parseFast()
###Always read the first line virtual / not
if line==0:
offsetAfterFirstImageRecord = fp.tell()
yr = imageData.metadata['Sensor acquisition year']
dys = imageData.metadata['Sensor acquisition day of year']
msecs = imageData.metadata['Sensor acquisition milliseconds of day']
usecs = imageData.metadata['Sensor acquisition micro-seconds of day']
self.sensingStart = datetime.datetime(yr,1,1) + datetime.timedelta(days=(dys-1)) + datetime.timedelta(seconds = usecs*1e-6)
self.nearRange = imageData.metadata['Slant range to 1st data sample']
self.prf = imageData.metadata['PRF'] * 1.0e-3
sceneCenterIncidenceAngle = self.parent.leaderFile.sceneHeaderRecord.metadata['Incidence angle at scene centre']
sarChannelId = imageData.metadata['SAR channel indicator']
scanId = imageData.metadata['Scan ID'] #Scan ID starts with 1
###Exit loop after first line if virtual
if virtual:
break
###Write line to file if not virtual
IQLine = np.fromfile(fp, dtype='>f', count=2*dataLen)
self.writeRawData(output, IQLine)
fp.close()
####If virtual file was requested, create VRT here
if virtual:
##Close input file
with open(output, 'w') as fid:
fid.write('''<VRTDataset rasterXSize="{0}" rasterYSize="{1}">
<VRTRasterBand dataType="CFloat32" band="1" subClass="VRTRawRasterBand">
<SourceFilename relativeToVRT="0">{2}</SourceFilename>
<ByteOrder>MSB</ByteOrder>
<ImageOffset>{3}</ImageOffset>
<PixelOffset>8</PixelOffset>
<LineOffset>{4}</LineOffset>
</VRTRasterBand>
</VRTDataset>'''.format(self.width, self.length,
os.path.abspath(self.file),
offsetAfterFirstImageRecord,
dataLen*8 + offsetAfterFirstImageRecord - offsetAfterImageFDR))
else:
##Close actual file on disk
output.close()
#burst parameters, currently only for the second, dual polarization, ScanSAR nominal mode
#that is the second WBD mode.
#p.25 and p.115 of ALOS-2/PALSAR-2 Level 1.1/1.5/2.1/3.1 CEOS SAR Product Format Description
#for the definations of wide swath mode
nbraw = [358, 470, 358, 355, 487]
ncraw = [2086.26, 2597.80, 1886.18, 1779.60, 2211.17]
self.parent.frame.nbraw = nbraw[scanId-1]
self.parent.frame.ncraw = ncraw[scanId-1]
#this is the prf fraction (total azimuth bandwith) used in extracting burst.
#here the total bandwith is 0.93 * prfs[3] for all subswaths, which is the following values:
#[0.7933, 0.6371, 0.8774, 0.9300, 0.7485]
prfs=[2661.847, 3314.512, 2406.568, 2270.575, 2821.225]
#Only needed for burst extraction. Skipping for now ....
#self.parent.frame.prffrac = 0.93 * prfs[3]/prfs[scanId-1]
self.sensingStop = self.sensingStart + datetime.timedelta(seconds = (self.length-1)/self.prf)
def _calculateRawDimensions(self,fp):
"""
Run through the data file once, and calculate the valid sampling window start time range.
"""
self.length = self.imageFDR.metadata['Number of SAR DATA records']
#self.width = self.imageFDR.metadata['SAR DATA record length']
self.width = self.imageFDR.metadata['Number of pixels per line per SAR channel']
return None