mirror/tesseract
1
0
Fork 0
mirror of https://github.com/tesseract-ocr/tesseract.git synced 2025-08-06 13:56:47 +08:00
Code Issues Actions 10 Packages Projects Releases Wiki Activity

Deleted lots of dead code, including PBLOB

Browse Source 
git-svn-id: https://tesseract-ocr.googlecode.com/svn/trunk@560 d0cd1f9f-072b-0410-8dd7-cf729c803f20
This commit is contained in:
theraysmith 2011-03-18 21:53:35 +00:00
parent df738bb9a4
commit 7cd3c74419
4 changed files with 9 additions and 50 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
dict/dict.cpp
View File

@ -75,21 +75,12 @@ Dict::Dict(Image* image_ptr)
getImage()->getCCUtil()->params()),
INT_MEMBER(hyphen_debug_level, 0, "Debug level for hyphenated words.",
getImage()->getCCUtil()->params()),
INT_MEMBER(ngram_permuter_debug_level, 0,
"Debug level for the ngram permuter.",
getImage()->getCCUtil()->params()),
double_MEMBER(ngram_permuter_nonmatch_score, -40.0,
"Average classifier score of a non-matching unichar.",
getImage()->getCCUtil()->params()),
INT_MEMBER(max_viterbi_list_size, 10, "Maximum size of viterbi list.",
getImage()->getCCUtil()->params()),
BOOL_MEMBER(use_only_first_uft8_step, false,
"Use only the first UTF8 step of the given string"
" when computing log probabilities.",
getImage()->getCCUtil()->params()),
double_MEMBER(ngram_model_scale_factor, 1.0, "Relative strength of the"
" ngram model relative to the character classifier ",
getImage()->getCCUtil()->params()),
double_MEMBER(certainty_scale, 20.0, "Certainty scaling factor",
getImage()->getCCUtil()->params()),
double_MEMBER(stopper_nondict_certainty_base, -2.50,

29
dict/dict.h
View File

@ -451,13 +451,6 @@ class Dict {
/// Returns true if the word looks like an absolute garbage
/// (e.g. image mistakenly recognized as text).
bool absolute_garbage(const WERD_CHOICE &word, const UNICHARSET &unicharset);
/* permngram.cpp ***********************************************************/
WERD_CHOICE *ngram_permute_and_select(
const BLOB_CHOICE_LIST_VECTOR &char_choices,
float rating_limit, float adjust_factor);
float compute_ngram_cost(float certainty, float denom,
const char *str, const char *context,
const UNICHARSET &unicharset);
/* dict.cpp ****************************************************************/
@ -532,8 +525,6 @@ class Dict {
int def_letter_is_okay(void* void_dawg_args,
UNICHAR_ID unichar_id, bool word_end);
int new_letter_is_okay(void* void_dawg_args,
UNICHAR_ID unichar_id, bool word_end);
int (Dict::*letter_is_okay_)(void* void_dawg_args,
UNICHAR_ID unichar_id, bool word_end);
/// Calls letter_is_okay_ member function.
@ -544,7 +535,8 @@ class Dict {
/// Probability in context function used by the ngram permuter.
double (Dict::*probability_in_context_)(const char* context,
double (Dict::*probability_in_context_)(const char* lang,
const char* context,
int context_bytes,
const char* character,
int character_bytes);
@ -553,13 +545,15 @@ class Dict {
int context_bytes,
const char* character,
int character_bytes) {
return (this->*probability_in_context_)(context, context_bytes,
character, character_bytes);
return (this->*probability_in_context_)(
getImage()->getCCUtil()->lang.string(),
context, context_bytes,
character, character_bytes);
}
/// Default (no-op) implementation of probability in context function.
double def_probability_in_context(
const char* context, int context_bytes,
const char* lang, const char* context, int context_bytes,
const char* character, int character_bytes) {
(void) context;
(void) context_bytes;
@ -567,7 +561,8 @@ class Dict {
(void) character_bytes;
return 0.0;
}
double ngram_probability_in_context(const char* context,
double ngram_probability_in_context(const char* lang,
const char* context,
int context_bytes,
const char* character,
int character_bytes);
@ -779,16 +774,10 @@ class Dict {
INT_VAR_H(dawg_debug_level, 0, "Set to 1 for general debug info"
", to 2 for more details, to 3 to see all the debug messages");
INT_VAR_H(hyphen_debug_level, 0, "Debug level for hyphenated words.");
INT_VAR_H(ngram_permuter_debug_level, 0,
"Debug level for the ngram permuter.");
double_VAR_H(ngram_permuter_nonmatch_score, -40.0,
"Average classifier score of a non-matching unichar.");
INT_VAR_H(max_viterbi_list_size, 10, "Maximum size of viterbi list.");
BOOL_VAR_H(use_only_first_uft8_step, false,
"Use only the first UTF8 step of the given string"
" when computing log probabilities.");
double_VAR_H(ngram_model_scale_factor, 1.0, "Relative strength of the"
" ngram model relative to the character classifier ");
double_VAR_H(certainty_scale, 20.0, "Certainty scaling factor");
double_VAR_H(stopper_nondict_certainty_base, -2.50,
"Certainty threshold for non-dict words");

18
dict/permute.cpp
View File

@ -49,7 +49,6 @@
#include "image.h"
#include "globals.h"
#include "ndminx.h"
#include "permngram.h"
#include "ratngs.h"
#include "stopper.h"
#include "tprintf.h"
@ -365,15 +364,6 @@ WERD_CHOICE *Dict::permute_all(const BLOB_CHOICE_LIST_VECTOR &char_choices,
LogNewChoice(adjust_factor, certainties, false, result2);
result1 = get_best_delete_other(result1, result2);
if (ngram_permuter_activated &&
(best_choice->rating() == WERD_CHOICE::kBadRating ||
best_choice->permuter() == TOP_CHOICE_PERM) &&
result1->permuter() == TOP_CHOICE_PERM) {
result2 = ngram_permute_and_select(char_choices, best_choice->rating(),
segment_reward_ngram_best_choice);
result1 = get_best_delete_other(result1, result2);
}
if (segment_segcost_rating) incorporate_segcost(result1);
} else {
result1 = permute_top_choice(char_choices, &top_choice_rating_limit,
@ -402,14 +392,6 @@ WERD_CHOICE *Dict::permute_all(const BLOB_CHOICE_LIST_VECTOR &char_choices,
result2 = permute_compound_words(char_choices, best_choice->rating());
result1 = get_best_delete_other(result1, result2);
if (ngram_permuter_activated &&
best_choice->permuter() < SYSTEM_DAWG_PERM &&
result1->permuter() < SYSTEM_DAWG_PERM) {
result2 = ngram_permute_and_select(char_choices, best_choice->rating(),
segment_penalty_ngram_best_choice);
result1 = get_best_delete_other(result1, result2);
}
}
return result1;
}

3
dict/permute.h
View File

@ -65,9 +65,6 @@ extern double_VAR_H(segment_reward_ngram_best_choice, 0.99,
"Score multipler for ngram permuter's best choice"
" (only used in the Han script path).");
extern BOOL_VAR_H(ngram_permuter_activated, false,
"Activate character-level n-gram-based permuter");
extern INT_VAR_H(max_permuter_attempts, 100000,
"Maximum number of different character choices to consider"
" during permutation. This limit is especially useful when"