| textblock.cc (ocrad-0.24) | : | textblock.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 29 | | skipping to change at line 29 |
|---|
| #include <climits> | | #include <climits> |
| #include <cstdio> | | #include <cstdio> |
| #include <string> | | #include <string> |
| #include <vector> | | #include <vector> |
| #include <stdint.h> | | #include <stdint.h> |
| | |
| #include "common.h" | | #include "common.h" |
| #include "rational.h" | | #include "rational.h" |
| #include "rectangle.h" | | #include "rectangle.h" |
| #include "track.h" | | #include "track.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" |
| #include "textblock.h" | | #include "textblock.h" |
| | |
| namespace { | | namespace { |
| | |
| void insert_line( std::vector< Textline * > & textlinep_vector, int i ) | | void insert_line( std::vector< Textline * > & textlinep_vector, int i ) |
| | |
| skipping to change at line 87 | | skipping to change at line 89 |
|---|
| joined = true; break; | | joined = true; break; |
| } | | } |
| } | | } |
| if( !joined ) ++i; | | if( !joined ) ++i; |
| } | | } |
| } | | } |
| } | | } |
| | |
| } // end namespace | | } // end namespace |
| | |
|
| | void Textblock::apply_filters( const Control & control ) |
| | { |
| | if( textlines() <= 0 ) return; |
| | for( unsigned f = 0; f < control.filters.size(); ++f ) |
| | { |
| | if( control.filters[f].user_filterp ) |
| | { |
| | for( int i = 0; i < textlines(); ++i ) |
| | tlpv[i]->apply_user_filter( *control.filters[f].user_filterp ); |
| | continue; |
| | } |
| | const Filter::Type filter = control.filters[f].type; |
| | if( filter != Filter::text_block ) |
| | { |
| | for( int i = 0; i < textlines(); ++i ) |
| | tlpv[i]->apply_filter( filter ); |
| | continue; |
| | } |
| | int l = right(), t = bottom(), r = left(), b = top(); |
| | for( int i = 0; i < textlines(); ++i ) |
| | { |
| | const Textline & line = *tlpv[i]; |
| | int first = line.characters(), last = -1; |
| | for( int j = line.big_initials(); j < line.characters(); ++j ) |
| | if( line.is_key_character( j ) ) |
| | { l = std::min( l, line.character(j).left() ); first = j; break; } |
| | for( int j = line.characters() - 1; j >= first; --j ) |
| | if( line.is_key_character( j ) ) |
| | { r = std::max( r, line.character(j).right() ); last = j; break; } |
| | if( i == 0 ) |
| | { for( int j = first; j <= last; ++j ) |
| | if( line.is_key_character( j ) ) |
| | t = std::min( t, line.character(j).top() ); } |
| | else if( i == textlines() - 1 ) |
| | { for( int j = first; j <= last; ++j ) |
| | if( line.is_key_character( j ) ) |
| | b = std::max( b, line.character(j).bottom() ); } |
| | } |
| | if( r < l || b < t ) continue; // can't apply filter; no text |
| | Rectangle re( l, t, r, b ); |
| | for( int i = 0; i < textlines(); ++i ) |
| | { |
| | Textline & line = *tlpv[i]; |
| | bool modified = false; |
| | for( int j = line.characters() - 1; j >= 0; --j ) |
| | if( line.character(j).height() >= 2 * line.height() || |
| | !re.includes( line.character(j).vcenter(), line.character(j).hcenter |
| | () ) ) |
| | { line.delete_character( j ); modified = true; } |
| | if( modified ) line.remove_leadind_trailing_duplicate_spaces(); |
| | } |
| | } |
| | } |
| | |
| Textblock::Textblock( const Rectangle & page, const Rectangle & block, | | Textblock::Textblock( const Rectangle & page, const Rectangle & block, |
| std::vector< Blob * > & blobp_vector ) | | std::vector< Blob * > & blobp_vector ) |
| : Rectangle( block ) | | : Rectangle( block ) |
| { | | { |
| std::vector< Blob * > pending; | | std::vector< Blob * > pending; |
| std::vector< Blob * > pending_tall; | | std::vector< Blob * > pending_tall; |
| std::vector< Blob * > pending_short; | | std::vector< Blob * > pending_short; |
| | |
| for( unsigned begin = 0, end = 0; end < blobp_vector.size(); begin = end ) | | for( unsigned begin = 0, end = 0; end < blobp_vector.size(); begin = end ) |
| { | | { |
| | |
| skipping to change at line 368 | | skipping to change at line 423 |
|---|
| { | | { |
| // Recognize characters. | | // Recognize characters. |
| for( int i = 0; i < textlines(); ++i ) | | for( int i = 0; i < textlines(); ++i ) |
| { | | { |
| // First pass. Recognize the easy characters. | | // First pass. Recognize the easy characters. |
| tlpv[i]->recognize1( control.charset ); | | tlpv[i]->recognize1( control.charset ); |
| // Second pass. Use context to clear up ambiguities. | | // Second pass. Use context to clear up ambiguities. |
| tlpv[i]->recognize2( control.charset ); | | tlpv[i]->recognize2( control.charset ); |
| } | | } |
| | |
|
| for( unsigned j = 0; j < control.filters.size(); ++j ) | | apply_filters( control ); |
| for( int i = 0; i < textlines(); ++i ) | | |
| tlpv[i]->apply_filter( control.filters[j] ); | | |
| | |
| // Remove unrecognized lines. | | // Remove unrecognized lines. |
| for( int i = textlines() - 1; i >= 0; --i ) | | for( int i = textlines() - 1; i >= 0; --i ) |
| { | | { |
| Textline & line1 = *tlpv[i]; | | Textline & line1 = *tlpv[i]; |
| bool recognized = false; | | bool recognized = false; |
| for( int j = 0 ; j < line1.characters(); ++j ) | | for( int j = 0 ; j < line1.characters(); ++j ) |
| { if( line1.character( j ).guesses() ) { recognized = true; break; } } | | { if( line1.character( j ).guesses() ) { recognized = true; break; } } |
| if( !recognized ) delete_line( tlpv, i ); | | if( !recognized ) delete_line( tlpv, i ); |
| } | | } |
| | | |
| End of changes. 4 change blocks. |
|---|
| 4 lines changed or deleted | | 58 lines changed or added |
|---|
"Fossies" - the Fresh Open Source Software Archive 