1. The band synthesis independent of a raster image
# Independent raster image a band synthesis Import OS Import numpy AS NP from OSGeo Import GDAL the os.chdir (R & lt ' E: \ Desktop file path \ GDAL \ osgeopy-Data \ osgeopy-Data \ Washington ' ) band1_fn = R & lt ' E: \ Desktop file path \ gdal \ osgeopy-data \ osgeopy -data \ the Landsat \ Washington \ p047r027_7t20000730_z10_nn10.tif ' band2_fn = R & lt ' E: \ Desktop file path \ gdal \ osgeopy-data \ osgeopy -data \ the Landsat \ Washington \ p047r027_7t20000730_z10_nn20.tif ' band3_fn = R & lt ' E: \ Desktop file path \ GDAL \ osgeopy-Data \ osgeopy-Data \ the Landsat \ Washington \ p047r027_7t20000730_z10_nn30.tif ' in_ds = gdal.Open(band1_fn) in_band = in_ds.GetRasterBand(1) gtiff_driver = gdal.GetDriverByName('GTiff') out_ds = gtiff_driver.Create('nat_color.tif', in_band.XSize, in_band.YSize, 3, in_band.DataType) out_ds.SetProjection(in_ds.GetProjection()) print(in_ds.GetGeoTransform()) out_ds.SetGeoTransform(in_ds.GetGeoTransform()) in_data = in_band.ReadAsArray() out_band = out_ds.GetRasterBand(3) out_band.WriteArray(in_data) in_ds =gdal.Open (band2_fn) out_band = out_ds.GetRasterBand (2 ) out_band.WriteArray ( in_ds.ReadAsArray () ) # in data set call ReadAsArray, if the data set has a plurality of bands, then to obtain a three-dimensional array # if the data set is a single-band, two-dimensional array is returned. The role of GetRasterBand just returned to query the band in_ds = gdal.Open (band3_fn) out_band = out_ds.GetRasterBand (1 ) out_band.WriteArray (in_ds.ReadAsArray ()) out_ds.FlushCache () # to ensure that data is written to disk, refresh the cache for I in Range (. 1,. 4 ): out_ds.GetRasterBand (I) .ComputeStatistics (False) # statistics are calculated for each band out_ds.BuildOverviews ( ' Average' , [2,4,8,16,32 ]) # Create overview view del out_ds # release dataset
2.band.ReadAsArray(xoff,yoff,win_xsize,win_ysize,buf_xsize,buf_ysize,buf_obj)
# XOFF column is the starting point, win_xsize number of columns is read, buf_xsize number of columns of the array output, and if the former is different resampling will, buf_obj storage array type # Data type conversion data = np.empty ((3, 6 ), = DTYPE a float) band.ReadAsArrray ( 14000,6000,6,3, buf_obj = Data) Data = band.ReadAsArrray (1400,600,6,3) .astype (a float)
3. The raster processing block
import os import numpy as np from osgeo import gdal os.chdir(r'E:\桌面文件保存路径\gdal\osgeopy-data\osgeopy-data\Landsat\Washington') in_ds = gdal.Open('') in_band = in_ds.GetRasterBand(1) xsize = in_band.XSize ysize = in_band.YSize block_xsize, block_ysize = in_band.GetBlockSize() nodata = in_band.GetNoDataValue() #获取nodata out_ds = in_ds.GetDriver().Create('dem_feet.tif', xsize, ysize, 1, in_band.DataType) out_ds.SetProjection(in_ds.GetProjection()) out_ds.SetGeoTransform(in_ds.GetGeoTransform()) out_band = out_ds.GetRasterBand(1) for x in range(0, xsize, block_xsize): if x+block_xsize < xsize: cols = block_xsize else: cols = xsize-x for y in range(0, ysize, block_ysize): if y+block_ysize < ysize: rows = block_ysize else: rows = ysize-y data = in_band.ReadAsArray(x, y, cols, rows) data = np.where(data == nodata, nodata, data*3.28084) out_band.WriteArray(data, x, y) out_band.FlushCache() out_band.SetNoDataValue(nodata) #将nodata排除在外 out_band.ComputeStatistics(False) out_ds.BuildOverviews('average', [2, 4, 8, 16, 32]) del out_ds
4.GeoTransform
# Using real-world coordinates # geotransform # origin 0,3 x \ y coordinate of coordinate reality, generally projected coordinate # 1,5-pixel width and height (the height of a negative value) # 2,4 x \ y rotation gt = ds. GetGeoTransform () # forward transform coordinates from image coordinates to real inv_gt = gdal.InvGeoTransform (gt) # inverse transform Qoffsets = gdal.ApplyGeoTransfrom (inv_gt, 465 200, 5.296 million ) #ApplyGeoTransform returns float Switch needs to be passed to ReadAsArray integer XOFF, Yoff = map (int, Qoffsets) # Python built-in map function for mapping the data do value = band.ReadAsArrray (XOFF, Yoff,. 1,. 1 ) [0, 0] data = band.ReadAsArray () XOFF, Yoff Map = (int, gdal.ApplyGeoTransform (inv_gt, 465.2 thousand, 5.296 million )) valueData = [Yoff, XOFF] # numpy array [row, column], and different gdal
5. subset of the saved image
# Extracting and storing the image subset Import OS from OSGeo Import GDAL vashon_ulx, vashon_uly = 532 000, 5.2626 million vashon_lrx, vashon_lry = 548.5 thousand, 5.2415 million the os.chdir (R & lt ' E: \ Desktop file path \ gdal \ osgeopy-data \ osgeopy- Data \ the Landsat \ Washington ' ) in_ds = gdal.Open ( ' nat_color.tif ' ) in_gt = in_ds.GetGeoTransform () inv_gt = gdal.InvGeoTransform (in_gt) IF gdal.VersionInfo () [0] == ' . 1 ' : IF inv_gt [0] ==. 1 : inv_gt = inv_gt[1] else: raise RuntimeError('Inverse geotransform failed') elif inv_gt is None: raise RuntimeError('Inverse geotransform failed') offsets_ul = gdal.ApplyGeoTransform(inv_gt, vashon_ulx, vashon_uly) offsets_lr = gdal.ApplyGeoTransform(inv_gt, vashon_lrx, vashon_lry) off_ulx, off_uly = map(int, offsets_ul) #取整得到行列值 off_lrx, off_lry = map(int, offsets_lr) rows Off_lry = - off_uly Columns = off_lrx - off_ulx gtiff_driver = gdal.GetDriverByName ( ' GTiff ' ) out_ds = gtiff_driver.Create ( ' vashon2.tif ' , Columns, rows,. 3 ) out_ds.SetProjection (in_ds.GetProjection ()) subset_ulx, subset_uly gdal.ApplyGeoTransform = (in_gt, off_ulx, off_uly) # herein as forward conversion is to prevent the original fall somewhere in the middle left corner pixel Print (subset_ulx) # calculated 531995.25 \ 5262624.75 original left corner, is transformed intermediate pixel does not fall. Print (subset_uly) out_gt = List (in_gt) #元组转列表，可修改 out_gt[0] = subset_ulx out_gt[3] = subset_uly out_ds.SetGeoTransform(out_gt) for i in range(1, 4): in_band = in_ds.GetRasterBand(i) out_band = out_ds.GetRasterBand(i) data = in_band.ReadAsArray(off_ulx, off_uly, columns, rows) out_band.WriteArray(data) out_ds.FlushCache() del out_ds
6. The image resampling
import os import numpy as np from osgeo import gdal os.chdir(r'') in_ds = gdal.Open(r'') in_band = in_ds.GetRasterBand(1) out_rows = in_band.YSize*2 out_columns = in_band.XSize*2 gtiff_driver = gdal.GetDriverByName('GTiff') out_ds = gtiff_driver.Create('band_resampled.tif', out_columns, out_rows) out_ds.SetProjection(in_ds.GetProjection()) geotransform = list(in_ds.GetGeoTransform()) geotransform[1] /= 2 geotransform[5] /= 2 out_ds.SetGeoTransform(geotransform) data = in_band.ReadAsArray(buf_xsize=out_columns, buf_ysize=out_rows) out_band = out_ds.GetRasterBand(1) out_band.WriteArray(data) out_band.FlushCache() out_band.ComputeStatistics(False) out_ds.BuildOverviews('average', [2, 4, 8, 16, 32, 64]) del out_ds #重采样为更大的元素 data = np.empty((2, 3), np.int) #2行3列 band.ReadAsArray(1400, 6000, 6, 4, buf_obj=data)
7. The readRaster \ WriteRaster read stored as a sequence of bytes
# ReadRaster \ WriteRaster stored as a sequence of bytes read ds.ReadRaster (1400, 6000, 2, 2, band_list = [. 1 ]) ds.WriteRaster ( 1400, 6000,. 6,. 4, Data, band_list = [. 1 ]) # byte sequence readRaster Import OS Import numpy AS NP from OSGeo Import GDAL in_ds = gdal.Open (R & lt ' nat_color.tif ' ) out_rows = int (in_ds.RasterYSize / 2) # DS read directly the number of ranks out_columns = int (in_ds. RasterXSize / 2 ) num_bands = in_ds.RasterCount # DS read band number gtiff_driver = gdal.GetDriverByName (' GTiff ' ) out_ds = gtiff_driver.Create ( ' nat_color_resampled.tif ' , out_columns, out_rows, num_bands) out_ds.SetProjection (in_ds.GetProjection ()) geotransform = List (in_ds.GetGeoTransform ()) geotransform [ . 1] = 2 * geotransform [ . 5] = 2 * out_ds.SetGeoTransform (geotransform) data = in_ds.ReadRaster (buf_xsize = out_columns, buf_ysize = out_rows) # using a smaller buffer to read and write data, resampling out_ds.WriteRaster (0, 0, out_columns, out_rows , data) # write data as the output data source out_ds.FlushCache () for I in range(num_bands): out_ds.GetRasterBand(i+1).ComputeStatistics(False) out_ds.BuildOverviews('average', [2, 4, 8, 16]) del out_ds
8. The sub-data sets HDF (modis)
# Subdatasets HDF, GDAL version needs to support HDF from OSGeo Import GDAL ds = gdal.Open (r ' E: \ Desktop file path \ gdal \ osgeopy-data \ osgeopy -data \ Modis \ MYD13Q1.A2014313.h20v11 .005.2014330092746.hdf ' ) subdatasets = ds.GetSubDatasets () # returns the list of sub data sets Print ( ' Number The subdatasets of:. {} the format (len (subdatasets)) ' ) for SD in subdatasets: Print ( ' the Name: {0 } \ nDescription: {}. 1 \ n- ' .format (* SD)) ndvi_ds = gdal.Open (subdatasets [0] [0]) # Of a [0] is read first set of data, the second data set name of the current is read
9.xml save images (driver.CreateCopy)
ds = gdal.Open('listing9—6.xml') gdal.GetDriverByName('PNG').CreateCopy('liberty.png', ds) #driver.CreateCopy('', ds)用来复制ds