Question

我想使用IDL程序在ENVI中空间分配LANDSAT照片。我有超过150个图像，我想要分组，所以我想以批处理模式运行程序（没有交互）。我知道如何手动完成，但是我会使用什么命令通过IDL代码中的lat / long坐标对图像进行空间子集化？

Answer 1

对于单个文件，这是一些灵感。您可以通过构建来为大量文件执行相同的操作一个文件名列表并循环遍历它。

; define the image to be opened (could be in a loop), I believe it can also be a tif, img...
img_file='path/to/image.hdr'
envi_open_file,img_file,r_fid=fid
if (fid eq -1) then begin
    print, 'Error when opening file ',img_file
    return
  endif

; let's define some coordinates
XMap=[-70.0580916, -70.5006694]
YMap=[-32.6030694, -32.9797194]
; now convert coordinates into pixel position:
; the transformation function uses the image geographic information:
ENVI_CONVERT_FILE_COORDINATES, FID, XF, YF, XMap, YMap
; we must consider integer. Think twice here, maybe you need to floor() or ceil()
XF=ROUND(XF)
YF=ROUND(YF)

; read the image
envi_file_query, fid, DIMS=DIMS, NB=NB, NL=NL, NS=NS
pos  = lindgen(nb)
; and store it in an array
image=fltarr(NS, NL, NB)
; read each band sequentially
FOR i=0, NB-1 DO BEGIN
   image[*,*,i]= envi_get_data(fid=fid, dims=dims, pos=pos[i])
endfor

; simply crop the data with array-indexing function
imagen= image[XF[0]:XF[1],YF[0]:YF[1]]

nl2=YF[1]-YF[0]
ns2=XF[1]-XF[0]

; read mapinfo to save it in the final file
map_info=envi_get_map_info(fid=fid)

envi_write_envi_file, imagen, data_type=4, $
  descrip = 'cropped', $
  map_info = map_info, $
  nl=nl2, ns=ns2, nb=nb, r_fid=r_fid, $
  OUT_NAME = 'path/to/cropped.hdr'

使用IDL和ENVI以批处理模式空间子集图像

1 个答案: