Hello, I'm trying to run an unsupervised classification (ISODATA) with the same mask image on jpg files in a folder through a loop and put the output files in another folder. I get the classification running ok, using a code for batch generating NDVI images that I found some time ago on the ITT website. I tried to modify this code to do the ISODATA classification that generates two classes and should apply a mask also. However, the outputs I get don't have the mask applied but instead the mask is loaded only into ENVI. There's also a few issues with the output file names, but that's not a big problem, as long as the output files have the mask applied. If there's anyone out there who can help me fix the code (see below), I'd appreciate it. Thanks, Mike pro ml_batch_01 compile_opt idl2 ;set up path to input and output files input_path = 'C:\B_Working\__ML_test\ml_in01\' ;change to output directory where I want the processed files to be placed cd, 'C:\B_Working\__ML_test\ml_out01\' ; search for data files in the specified directory files = FILE_SEARCH(input_path + '*.jpg', count=count) IF (count EQ 0) THEN BEGIN PrintF, 'No files were found to process' ENDIF for i=0, count-1 do begin envi_open_file, files[i], r_fid=fid if (fid eq -1) then begin envi_batch_exit return endif envi_file_query, fid, dims=dims, ns=ns, nl=nl, fname=filename ;; ML: setting up the mask file maskfile = 'C:\B_Working\Niina_data_16May2011\__ML_test\temp\ml_mask42' ENVI_OPEN_FILE, maskfile, r_fid=m_fid if (m_fid eq -1) then return envi_file_query, m_fid, dims=dims, nb=nb ;set pos array for calculating ISODATA (CLASS_DOIT method=4) pos = [1,2,3] - 1 ;set the output file names by stripping out the base name and appending '_unsup.tif' out_name = file_basename(filename, '.jpg') + '_unsup.tif' ;call the doit for the unsupervised ISODATA classification envi_doit, 'class_doit', $ fid=fid, pos=pos, dims=dims, $ /check, o_min=0, o_max=255, $ out_name=out_name, r_fid=r_fid, mfid=mfid, mpos=mpos, value=0, $ method=4, change_thresh=0.05, iso_merge_dist=5, $ ; method 4 is isodata iso_min_pixels=10, min_classes=2, max_classes=2, num_classes=2, iterations=5, $ iso_split_std=1.0, iso_merge_pairs=2 endfor end