despeckle before OCR.

Author: jashkenas

lib/docsplit/text_extractor.rb

@@ -63,14 +63,14 @@ def extract_from_ocr(pdf, pages)
         pages.each do |page|
           tiff = "#{tempdir}/#{@pdf_name}_#{page}.tif"
           file = "#{base_path}_#{page}"
-          run "MAGICK_TMPDIR=#{tempdir} OMP_NUM_THREADS=2 gm convert +adjoin #{MEMORY_ARGS} #{OCR_FLAGS} #{pdf}[#{page - 1}] #{tiff} 2>&1"
+          run "MAGICK_TMPDIR=#{tempdir} OMP_NUM_THREADS=2 gm convert -despeckle +adjoin #{MEMORY_ARGS} #{OCR_FLAGS} #{pdf}[#{page - 1}] #{tiff} 2>&1"
           run "tesseract #{tiff} #{file} -l eng 2>&1"
           clean_text(file + '.txt') if @clean_ocr
           FileUtils.remove_entry_secure tiff
         end
       else
         tiff = "#{tempdir}/#{@pdf_name}.tif"
-        run "MAGICK_TMPDIR=#{tempdir} OMP_NUM_THREADS=2 gm convert #{MEMORY_ARGS} #{OCR_FLAGS} #{pdf} #{tiff} 2>&1"
+        run "MAGICK_TMPDIR=#{tempdir} OMP_NUM_THREADS=2 gm convert -despeckle #{MEMORY_ARGS} #{OCR_FLAGS} #{pdf} #{tiff} 2>&1"
         run "tesseract #{tiff} #{base_path} -l eng 2>&1"
         clean_text(base_path + '.txt') if @clean_ocr
       end