Major internationalization improvements

git-svn-id: https://tesseract-ocr.googlecode.com/svn/trunk@134 d0cd1f9f-072b-0410-8dd7-cf729c803f20
allen8807 · Feb 1, 2008 · e8787f3 · e8787f3
1 parent 2a67830
commit e8787f3
Show file tree

Hide file tree

Showing 3 changed files with 7 additions and 1 deletion.
diff --git a/tessdata/configs/box.train b/tessdata/configs/box.train
@@ -14,3 +14,5 @@ edges_childarea 0.65
 edges_boxarea 0.9
 tessedit_resegment_from_boxes T
 tessedit_train_from_boxes T
+textord_repeat_extraction F
+textord_no_rejects T
diff --git a/tessdata/configs/kannada b/tessdata/configs/kannada
@@ -0,0 +1,4 @@
+textord_skewsmooth_offset 8
+textord_skewsmooth_offset2 8
+textord_merge_desc 0.5
+textord_no_rejects 1
diff --git a/training/wordlist2dawg.cpp b/training/wordlist2dawg.cpp
@@ -38,7 +38,7 @@ int main(int argc, char** argv) {
   const char* dawg_filename = argv[2];
 
   EDGE_ARRAY dawg;
-  INT32 max_num_edges =  20000000;
+  INT32 max_num_edges =  100000000;
   INT32 reserved_edges =  1000000;
 
   dawg = (EDGE_ARRAY) Emalloc(sizeof (EDGE_RECORD) * max_num_edges);