00001
00020 #ifndef CONTROL_H
00021 #define CONTROL_H
00022
00023 #include "varable.h"
00024 #include "ocrblock.h"
00025
00026 #include "ratngs.h"
00027 #include "statistc.h"
00028
00029 #include "ocrshell.h"
00030 #include "pageres.h"
00031 #include "charsample.h"
00032 #include "notdll.h"
00033
00034
00039 enum ACCEPTABLE_WERD_TYPE
00040 {
00041 AC_UNACCEPTABLE,
00042 AC_LOWER_CASE,
00043 AC_UPPER_CASE,
00044 AC_INITIAL_CAP,
00045 AC_LC_ABBREV,
00046 AC_UC_ABBREV
00047 };
00048
00049 typedef BOOL8 (*BLOB_REJECTOR) (PBLOB *, BLOB_CHOICE_IT *, void *);
00050
00051 extern INT_VAR_H (tessedit_single_match, FALSE, "Top choice only from CP");
00052
00053 extern BOOL_VAR_H (tessedit_print_text, FALSE, "Write text to stdout");
00054 extern BOOL_VAR_H (tessedit_draw_words, FALSE, "Draw source words");
00055 extern BOOL_VAR_H (tessedit_draw_outwords, FALSE, "Draw output words");
00056 extern BOOL_VAR_H (tessedit_training_wiseowl, FALSE, "Call WO to learn blobs");
00057 extern BOOL_VAR_H (tessedit_training_tess, FALSE, "Call Tess to learn blobs");
00058 extern BOOL_VAR_H (tessedit_matcher_is_wiseowl, FALSE, "Call WO to classify");
00059 extern BOOL_VAR_H (tessedit_dump_choices, FALSE, "Dump char choices");
00060 extern BOOL_VAR_H (tessedit_fix_fuzzy_spaces, TRUE,
00061 "Try to improve fuzzy spaces");
00062 extern BOOL_VAR_H (tessedit_unrej_any_wd, FALSE,
00063 "Dont bother with word plausibility");
00064 extern BOOL_VAR_H (tessedit_fix_hyphens, TRUE, "Crunch double hyphens?");
00065 extern BOOL_VAR_H (tessedit_reject_fullstops, FALSE, "Reject all fullstops");
00066 extern BOOL_VAR_H (tessedit_reject_suspect_fullstops, FALSE,
00067 "Reject suspect fullstops");
00068 extern BOOL_VAR_H (tessedit_redo_xheight, TRUE, "Check/Correct x-height");
00069 extern BOOL_VAR_H (tessedit_cluster_adaption_on, TRUE,
00070 "Do our own adaption - ems only");
00071 extern BOOL_VAR_H (tessedit_enable_doc_dict, TRUE,
00072 "Add words to the document dictionary");
00073 extern BOOL_VAR_H (word_occ_first, FALSE, "Do word occ before re-est xht");
00074 extern BOOL_VAR_H (tessedit_xht_fiddles_on_done_wds, TRUE,
00075 "Apply xht fix up even if done");
00076 extern BOOL_VAR_H (tessedit_xht_fiddles_on_no_rej_wds, TRUE,
00077 "Apply xht fix up even in no rejects");
00078 extern INT_VAR_H (x_ht_check_word_occ, 2, "Check Char Block occupancy");
00079 extern INT_VAR_H (x_ht_stringency, 1, "How many confirmed a/n to accept?");
00080 extern BOOL_VAR_H (x_ht_quality_check, TRUE, "Dont allow worse quality");
00081 extern BOOL_VAR_H (tessedit_debug_block_rejection, FALSE, "Block and Row stats");
00082 extern INT_VAR_H (debug_x_ht_level, 0, "Reestimate debug");
00083 extern BOOL_VAR_H (rej_use_xht, TRUE, "Individual rejection control");
00084 extern BOOL_VAR_H (debug_acceptable_wds, FALSE, "Dump word pass/fail chk");
00085 extern STRING_VAR_H (chs_leading_punct, "('`\"", "Leading punctuation");
00086 extern STRING_VAR_H (chs_trailing_punct1, ").,;:?!", "1st Trailing punctuation");
00087 extern STRING_VAR_H (chs_trailing_punct2, ")'`\"", "2nd Trailing punctuation");
00088 extern double_VAR_H (quality_rej_pc, 0.08,
00089 "good_quality_doc lte rejection limit");
00090 extern double_VAR_H (quality_blob_pc, 0.0,
00091 "good_quality_doc gte good blobs limit");
00092 extern double_VAR_H (quality_outline_pc, 1.0,
00093 "good_quality_doc lte outline error limit");
00094 extern double_VAR_H (quality_char_pc, 0.95,
00095 "good_quality_doc gte good char limit");
00096 extern INT_VAR_H (quality_min_initial_alphas_reqd, 2,
00097 "alphas in a good word");
00098 extern BOOL_VAR_H (tessedit_tess_adapt_to_rejmap, FALSE,
00099 "Use reject map to control Tesseract adaption");
00100 extern INT_VAR_H (tessedit_tess_adaption_mode, 3,
00101 "Adaptation decision algorithm for tess");
00102 extern INT_VAR_H (tessedit_em_adaption_mode, 62,
00103 "Adaptation decision algorithm for ems matrix matcher");
00104 extern BOOL_VAR_H (tessedit_cluster_adapt_after_pass1, FALSE,
00105 "Adapt using clusterer after pass 1");
00106 extern BOOL_VAR_H (tessedit_cluster_adapt_after_pass2, FALSE,
00107 "Adapt using clusterer after pass 1");
00108 extern BOOL_VAR_H (tessedit_cluster_adapt_after_pass3, FALSE,
00109 "Adapt using clusterer after pass 1");
00110 extern BOOL_VAR_H (tessedit_cluster_adapt_before_pass1, FALSE,
00111 "Adapt using clusterer before Tess adaping during pass 1");
00112 extern INT_VAR_H (tessedit_cluster_adaption_mode, 0,
00113 "Adaptation decision algorithm for matrix matcher");
00114 extern BOOL_VAR_H (tessedit_adaption_debug, FALSE,
00115 "Generate and print debug information for adaption");
00116 extern BOOL_VAR_H (tessedit_minimal_rej_pass1, FALSE,
00117 "Do minimal rejection on pass 1 output");
00118 extern BOOL_VAR_H (tessedit_test_adaption, FALSE,
00119 "Test adaption criteria");
00120 extern BOOL_VAR_H (tessedit_global_adaption, FALSE,
00121 "Adapt to all docs over time");
00122 extern BOOL_VAR_H (tessedit_matcher_log, FALSE, "Log matcher activity");
00123 extern INT_VAR_H (tessedit_test_adaption_mode, 3,
00124 "Adaptation decision algorithm for tess");
00125 extern BOOL_VAR_H (test_pt, FALSE, "Test for point");
00126 extern double_VAR_H (test_pt_x, 99999.99, "xcoord");
00127 extern double_VAR_H (test_pt_y, 99999.99, "ycoord");
00128 void recog_pseudo_word(
00129 BLOCK_LIST *block_list,
00130 BOX &selection_box);
00131 BOOL8 recog_interactive(
00132 BLOCK *,
00133 ROW *row,
00134 WERD *word
00135 );
00136 void recog_all_words(
00137 PAGE_RES *page_res,
00138 volatile ETEXT_DESC *monitor
00139 );
00140 void classify_word_pass1(
00141 WERD_RES *word,
00142 ROW *row,
00143 BOOL8 cluster_adapt,
00144 CHAR_SAMPLES_LIST *char_clusters,
00145 CHAR_SAMPLE_LIST *chars_waiting);
00146
00147 void classify_word_pass2(WERD_RES *word, ROW *row);
00148 void match_word_pass2(
00149 WERD_RES *word,
00150 ROW *row,
00151 float x_height);
00152 void fix_rep_char(
00153 WERD_RES *word
00154 );
00155 void fix_quotes(
00156 char *string,
00157 WERD *word,
00158 BLOB_CHOICE_LIST_CLIST *blob_choices);
00159 void fix_hyphens(
00160 char *string,
00161 WERD *word,
00162 BLOB_CHOICE_LIST_CLIST *blob_choices);
00163 void merge_blobs(
00164 PBLOB *blob1,
00165 PBLOB *blob2
00166 );
00167 void choice_dump_tester(
00168 PBLOB *,
00169 DENORM *,
00170 BOOL8 correct,
00171 char *text,
00172 INT32 count,
00173 BLOB_CHOICE_LIST *ratings
00174 );
00175 WERD *make_bln_copy(WERD *src_word, ROW *row, float x_height, DENORM *denorm);
00176 ACCEPTABLE_WERD_TYPE acceptable_word_string(const char *s);
00177 BOOL8 check_debug_pt(WERD_RES *word, int location);
00178 void set_word_fonts(
00179 WERD_RES *word,
00180 BLOB_CHOICE_LIST_CLIST *blob_choices);
00181 void font_recognition_pass(
00182 PAGE_RES_IT &page_res_it);
00183 void add_in_one_row(
00184 ROW_RES *row,
00185 STATS *fonts,
00186 INT8 *italic,
00187 INT8 *bold
00188 );
00189 void find_modal_font(
00190 STATS *fonts,
00191 INT8 *font_out,
00192 INT8 *font_count
00193 );
00194 #endif