| textline.cc (ocrad-0.24) | : | textline.cc (ocrad-0.25) |
|---|
| /* GNU Ocrad - Optical Character Recognition program | | /* GNU Ocrad - Optical Character Recognition program |
|
| Copyright (C) 2003-2014 Antonio Diaz Diaz. | | Copyright (C) 2003-2015 Antonio Diaz Diaz. |
| | |
| This program is free software: you can redistribute it and/or modify | | This program is free software: you can redistribute it and/or modify |
| it under the terms of the GNU General Public License as published by | | it under the terms of the GNU General Public License as published by |
| the Free Software Foundation, either version 2 of the License, or | | the Free Software Foundation, either version 2 of the License, or |
| (at your option) any later version. | | (at your option) any later version. |
| | |
| This program is distributed in the hope that it will be useful, | | This program is distributed in the hope that it will be useful, |
| but WITHOUT ANY WARRANTY; without even the implied warranty of | | but WITHOUT ANY WARRANTY; without even the implied warranty of |
| MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| GNU General Public License for more details. | | GNU General Public License for more details. |
| | |
| skipping to change at line 32 | | skipping to change at line 32 |
|---|
| #include <string> | | #include <string> |
| #include <vector> | | #include <vector> |
| #include <stdint.h> | | #include <stdint.h> |
| | |
| #include "common.h" | | #include "common.h" |
| #include "histogram.h" | | #include "histogram.h" |
| #include "rational.h" | | #include "rational.h" |
| #include "rectangle.h" | | #include "rectangle.h" |
| #include "track.h" | | #include "track.h" |
| #include "ucs.h" | | #include "ucs.h" |
|
| | #include "user_filter.h" |
| #include "bitmap.h" | | #include "bitmap.h" |
| #include "blob.h" | | #include "blob.h" |
| #include "character.h" | | #include "character.h" |
| #include "page_image.h" | | #include "page_image.h" |
| #include "textline.h" | | #include "textline.h" |
| | |
| namespace { | | namespace { |
| | |
|
| // Return the character position >= first preceding a big gap or eol. | | // Returns the character position >= first preceding a big gap or eol. |
| // | | // |
| int find_big_gap( const Textline & line, const int first, | | int find_big_gap( const Textline & line, const int first, |
| const int space_width_limit ) | | const int space_width_limit ) |
| { | | { |
| int i = first; | | int i = first; |
| while( i + 1 < line.characters() ) | | while( i + 1 < line.characters() ) |
| { | | { |
| const Character & c1 = line.character( i ); | | const Character & c1 = line.character( i ); |
| const Character & c2 = line.character( i + 1 ); | | const Character & c2 = line.character( i + 1 ); |
| const int gap = c2.left() - c1.right() - 1; | | const int gap = c2.left() - c1.right() - 1; |
| | |
| skipping to change at line 120 | | skipping to change at line 121 |
|---|
| for( int i = 0; i < characters(); ++i ) | | for( int i = 0; i < characters(); ++i ) |
| if( cpv[i]->h_includes( col ) ) return cpv[i]; | | if( cpv[i]->h_includes( col ) ) return cpv[i]; |
| return 0; | | return 0; |
| } | | } |
| | |
| Rectangle Textline::charbox( const Character & c ) const | | Rectangle Textline::charbox( const Character & c ) const |
| { | | { |
| return Rectangle( c.left(), top( c.hcenter() ), c.right(), bottom( c.hcenter()
) ); | | return Rectangle( c.left(), top( c.hcenter() ), c.right(), bottom( c.hcenter()
) ); |
| } | | } |
| | |
|
| | bool Textline::is_key_character( const int i ) const |
| | { |
| | if( i < big_initials_ || i >= characters() ) |
| | Ocrad::internal_error( "is_key_character, index out of bounds." ); |
| | return ( cpv[i]->isalnum() && cpv[i]->guess( 0 ).code != 'J' && |
| | cpv[i]->height() < 2 * height() && 2 * cpv[i]->height() > height() ); |
| | } |
| | |
| void Textline::delete_character( const int i ) | | void Textline::delete_character( const int i ) |
| { | | { |
| if( i < 0 || i >= characters() ) | | if( i < 0 || i >= characters() ) |
| Ocrad::internal_error( "delete_character, index out of bounds." ); | | Ocrad::internal_error( "delete_character, index out of bounds." ); |
| if( i < big_initials_ ) --big_initials_; | | if( i < big_initials_ ) --big_initials_; |
| delete cpv[i]; cpv.erase( cpv.begin() + i ); | | delete cpv[i]; cpv.erase( cpv.begin() + i ); |
| } | | } |
| | |
| int Textline::shift_characterp( Character * const p, const bool big ) | | int Textline::shift_characterp( Character * const p, const bool big ) |
| { | | { |
| | |
| skipping to change at line 279 | | skipping to change at line 288 |
|---|
| | |
| void Textline::dprint( const Control & control, const bool graph, | | void Textline::dprint( const Control & control, const bool graph, |
| const bool recursive ) const | | const bool recursive ) const |
| { | | { |
| if( graph || recursive ) | | if( graph || recursive ) |
| { | | { |
| Histogram hist; | | Histogram hist; |
| for( int i = 0; i < characters(); ++i ) | | for( int i = 0; i < characters(); ++i ) |
| if( !character(i).maybe(' ') ) | | if( !character(i).maybe(' ') ) |
| hist.add_sample( character(i).height() ); | | hist.add_sample( character(i).height() ); |
|
| std::fprintf( control.outfile, "mean height = %d, median height = %d, track | | std::fprintf( control.outfile, "mean height = %d, median height = %d, track |
| segments = %d\n", | | segments = %d, big initials = %d\n", |
| mean_height(), hist.median(), segments() ); | | mean_height(), hist.median(), segments(), big_initials_ ); |
| } | | } |
| | |
| for( int i = 0; i < characters(); ++i ) | | for( int i = 0; i < characters(); ++i ) |
| { | | { |
| const Character & c = character( i ); | | const Character & c = character( i ); |
| if( i < big_initials_ ) c.dprint( control, c, graph, recursive ); | | if( i < big_initials_ ) c.dprint( control, c, graph, recursive ); |
| else c.dprint( control, charbox( c ), graph, recursive ); | | else c.dprint( control, charbox( c ), graph, recursive ); |
| } | | } |
| std::fputs( "\n", control.outfile ); | | std::fputs( "\n", control.outfile ); |
| } | | } |
| | |
| skipping to change at line 326 | | skipping to change at line 335 |
|---|
| c.only_guess( UCS::toupper( code ), 0 ); | | c.only_guess( UCS::toupper( code ), 0 ); |
| } | | } |
| } | | } |
| else c.recognize1( charset, charbox( c ) ); | | else c.recognize1( charset, charbox( c ) ); |
| } | | } |
| } | | } |
| | |
| void Textline::apply_filter( const Filter::Type filter ) | | void Textline::apply_filter( const Filter::Type filter ) |
| { | | { |
| bool modified = false; | | bool modified = false; |
|
| for( int i = characters() - 1; i >= 0; --i ) | | |
| { | | |
| Character & c = character( i ); | | |
| if( !c.guesses() ) continue; | | |
| c.apply_filter( filter ); | | |
| if( !c.guesses() ) { delete_character( i ); modified = true; } | | |
| } | | |
| if( filter == Filter::same_height ) | | if( filter == Filter::same_height ) |
| { | | { |
| Histogram hist; | | Histogram hist; |
| for( int i = 0; i < characters(); ++i ) | | for( int i = 0; i < characters(); ++i ) |
| if( !character(i).maybe(' ') ) | | if( !character(i).maybe(' ') ) |
| hist.add_sample( character(i).height() ); | | hist.add_sample( character(i).height() ); |
| const int median_height = hist.median(); | | const int median_height = hist.median(); |
| for( int i = characters() - 1; i >= 0; --i ) | | for( int i = characters() - 1; i >= 0; --i ) |
| if( !character(i).maybe(' ') && | | if( !character(i).maybe(' ') && |
|
| !Ocrad::similar( character(i).height(), median_height, 13, 2 ) ) | | !Ocrad::similar( character(i).height(), median_height, 10, 2 ) ) |
| { delete_character( i ); modified = true; } | | { delete_character( i ); modified = true; } |
| } | | } |
|
| if( modified ) // remove leadind/trailing/duplicate spaces | | else |
| | { |
| for( int i = characters() - 1; i >= 0; --i ) | | for( int i = characters() - 1; i >= 0; --i ) |
|
| if( character(i).maybe(' ') && | | { |
| ( i == 0 || i == characters() - 1 || character(i-1).maybe(' ') ) ) | | Character & c = character( i ); |
| delete_character( i ); | | if( !c.guesses() ) continue; |
| | c.apply_filter( filter ); |
| | if( !c.guesses() && filter != Filter::upper_num_mark ) |
| | { delete_character( i ); modified = true; } |
| | } |
| | if( filter == Filter::upper_num_mark ) |
| | join_broken_unrecognized_characters(); |
| | } |
| | if( modified ) remove_leadind_trailing_duplicate_spaces(); |
| | } |
| | |
| | void Textline::apply_user_filter( const User_filter & user_filter ) |
| | { |
| | bool modified = false; |
| | for( int i = characters() - 1; i >= 0; --i ) |
| | { |
| | Character & c = character( i ); |
| | if( !c.guesses() ) continue; |
| | c.apply_user_filter( user_filter ); |
| | if( !c.guesses() && user_filter.discard() ) |
| | { delete_character( i ); modified = true; } |
| | } |
| | if( user_filter.mark() ) join_broken_unrecognized_characters(); |
| | if( modified ) remove_leadind_trailing_duplicate_spaces(); |
| | } |
| | |
| | void Textline::join_broken_unrecognized_characters() |
| | { |
| | for( int i = characters() - 1; i > 0; --i ) |
| | if( !character(i).guesses() && |
| | character(i).h_overlaps( character( i - 1 ) ) ) |
| | delete_character( i ); |
| | } |
| | |
| | void Textline::remove_leadind_trailing_duplicate_spaces() |
| | { |
| | for( int i = characters() - 1; i >= 0; --i ) |
| | if( character(i).maybe(' ') && |
| | ( i == 0 || i == characters() - 1 || character(i-1).maybe(' ') ) ) |
| | delete_character( i ); |
| } | | } |
| | | |
| End of changes. 9 change blocks. |
|---|
| 17 lines changed or deleted | | 59 lines changed or added |
|---|
"Fossies" - the Fresh Open Source Software Archive 