ENVI5.1 与ENVI5.3 ReprojectGLT结果差异

对风云3（FY3）条带亮度温度数据进行几何校正，结果发现相同代码下，ENVI5.1+IDL8.3与 ENVI5.3+IDL8.5运行结果明显不同

 

代码如下：

pro fy3d_geo

  in_path ='D:\_User\test_he5\'

  out_path_rad = 'D:\_User\test_rad\'

  out_path_geo = 'D:\_User\test_geo\'

  file_arr = file_search(in_path, '*.he5')

  fileCount = n_elements(file_arr)

  for i=0, fileCount-1 do begin

    print, 'Processing... ' + file_arr[i]

    file = file_arr[i]

    filename = file_basename(file, '.he5')

    hd_id = h5f_open(file)

    sd_id1 = h5d_open(hd_id,'/Geolocation/Latitude')

    sd_id2 = h5d_open(hd_id,'/Geolocation/Longitude')

    lat = h5d_read(sd_id1)

    lon = h5d_read(sd_id2)

    lat[where(lat eq 65535)] = 90

    lon[where(lon eq 65535)] = -180

    sd_id3 = h5d_open(hd_id,'/Calibration/EARTH_OBSERVE_BT_10_to_89GHz')

    bt = h5d_read(sd_id3)

    bt89V = bt[*,*,8]

    bt89V_rad = bt89V*0.01+327.68

    fileLat = out_path_rad+filename+'_lat.tif'

    fileLon = out_path_rad+filename+'_lon.tif'

    file89V = out_path_rad+filename+'_89V_rad.tif'

    write_tiff, fileLat, lat, /float

    write_tiff, fileLon, lon, /float

    write_tiff, file89V, bt89V_rad, /float

    e = envi(/headless)

    rasterLat = e.openraster(fileLat)

    rasterLon = e.openraster(fileLon)

    raster89V = e.openraster(file89V)

    ; 89V

    file_geo89V = out_path_geo+filename+'_89V_geo'

    Task = envitask('ReprojectGLT')

    Task.INPUT_RASTER = raster89V

    Task.LATITUDE_RASTER = rasterLat

    Task.LONGITUDE_RASTER = rasterLon

    Task.OUTPUT_RASTER_URI = file_geo89V

    Task.execute

    envi_open_file, file_geo89V, R_FID =FID

    envi_file_query, fid, dims=dims, nb=nb, fname=fname, data_type=data_type, interleave=interleave

    out_name = out_path_geo+filename+'_89V_geo.tif'

    envi_output_to_external_format, OUT_NAME = out_name, FID = fid, DIMS=dims, POS = indgen(nb), /TIFF

  endfor

  print, 'finish'

end

ENVI5.3+IDL8.5结果如下

 

ENVI5.1+IDL8.3结果如下

 

ENVI5.1+IDL8.3的ReprojectGLT的结果在质量上明显优于ENVi5.3+IDL8.5