tesseract::ImageData Class Reference

#include <imagedata.h>

Public Member Functions
	ImageData ()
	ImageData (bool vertical, Pix *pix)
	~ImageData ()
bool	Serialize (TFile *fp) const
bool	DeSerialize (bool swap, TFile *fp)
const STRING &	imagefilename () const
void	set_imagefilename (const STRING &name)
int	page_number () const
void	set_page_number (int num)
const GenericVector< char > &	image_data () const
const STRING &	language () const
void	set_language (const STRING &lang)
const STRING &	transcription () const
const GenericVector< TBOX > &	boxes () const
const GenericVector< STRING > &	box_texts () const
const STRING &	box_text (int index) const
void	SetPix (Pix *pix)
Pix *	GetPix () const
Pix *	PreScale (int target_height, int max_height, float *scale_factor, int *scaled_width, int *scaled_height, GenericVector< TBOX > *boxes) const
int	MemoryUsed () const
void	Display () const
void	AddBoxes (const GenericVector< TBOX > &boxes, const GenericVector< STRING > &texts, const GenericVector< int > &box_pages)

Static Public Member Functions
static ImageData *	Build (const char *name, int page_number, const char *lang, const char *imagedata, int imagedatasize, const char *truth_text, const char *box_text)
static bool	SkipDeSerialize (bool swap, tesseract::TFile *fp)

Detailed Description

Definition at line 103 of file imagedata.h.

Constructor & Destructor Documentation

ImageData() [1/2]

tesseract::ImageData::ImageData

(

)

Definition at line 122 of file imagedata.cpp.

                     : page_number_(-1), vertical_text_(false) {
}

ImageData() [2/2]

tesseract::ImageData::ImageData	(	bool	vertical,
		Pix *	pix
	)

Definition at line 125 of file imagedata.cpp.

  : page_number_(0), vertical_text_(vertical) {
  SetPix(pix);
}

~ImageData()

tesseract::ImageData::~ImageData

(

)

Definition at line 129 of file imagedata.cpp.

                      {
}

Member Function Documentation

AddBoxes()

void tesseract::ImageData::AddBoxes	(	const GenericVector< TBOX > &	boxes,
		const GenericVector< STRING > &	texts,
		const GenericVector< int > &	box_pages
	)

Definition at line 311 of file imagedata.cpp.

                                                              {
  // Copy the boxes and make the transcription.
  for (int i = 0; i < box_pages.size(); ++i) {
    if (page_number_ >= 0 && box_pages[i] != page_number_) continue;
    transcription_ += texts[i];
    boxes_.push_back(boxes[i]);
    box_texts_.push_back(texts[i]);
  }
}

box_text()

const STRING& tesseract::ImageData::box_text ( int  index ) const

inline

Definition at line 155 of file imagedata.h.

                                          {
    return box_texts_[index];
  }

box_texts()

const GenericVector<STRING>& tesseract::ImageData::box_texts ( ) const

inline

Definition at line 152 of file imagedata.h.

                                                 {
    return box_texts_;
  }

boxes()

const GenericVector<TBOX>& tesseract::ImageData::boxes ( ) const

inline

Definition at line 149 of file imagedata.h.

                                           {
    return boxes_;
  }

Build()

ImageData * tesseract::ImageData::Build ( const char *  name, int  page_number, const char *  lang, const char *  imagedata, int  imagedatasize, const char *  truth_text, const char *  box_text  )

static

Definition at line 134 of file imagedata.cpp.

                                                                          {
  ImageData* image_data = new ImageData();
  image_data->imagefilename_ = name;
  image_data->page_number_ = page_number;
  image_data->language_ = lang;
  // Save the imagedata.
  image_data->image_data_.resize_no_init(imagedatasize);
  memcpy(&image_data->image_data_[0], imagedata, imagedatasize);
  if (!image_data->AddBoxes(box_text)) {
    if (truth_text == NULL || truth_text[0] == '\0') {
      tprintf("Error: No text corresponding to page %d from image %s!\n",
              page_number, name);
      delete image_data;
      return NULL;
    }
    image_data->transcription_ = truth_text;
    // If we have no boxes, the transcription is in the 0th box_texts_.
    image_data->box_texts_.push_back(truth_text);
    // We will create a box for the whole image on PreScale, to save unpacking
    // the image now.
  } else if (truth_text != NULL && truth_text[0] != '\0' &&
             image_data->transcription_ != truth_text) {
    // Save the truth text as it is present and disagrees with the box text.
    image_data->transcription_ = truth_text;
  }
  return image_data;
}

DeSerialize()

bool tesseract::ImageData::DeSerialize	(	bool	swap,
		TFile *	fp
	)

Definition at line 180 of file imagedata.cpp.

                                                {
  if (!imagefilename_.DeSerialize(swap, fp)) return false;
  if (fp->FRead(&page_number_, sizeof(page_number_), 1) != 1) return false;
  if (swap) ReverseN(&page_number_, sizeof(page_number_));
  if (!image_data_.DeSerialize(swap, fp)) return false;
  if (!transcription_.DeSerialize(swap, fp)) return false;
  // WARNING: Will not work across different endian machines.
  if (!boxes_.DeSerialize(swap, fp)) return false;
  if (!box_texts_.DeSerializeClasses(swap, fp)) return false;
  inT8 vertical = 0;
  if (fp->FRead(&vertical, sizeof(vertical), 1) != 1) return false;
  vertical_text_ = vertical != 0;
  return true;
}

Display()

void tesseract::ImageData::Display

(

)

const

Definition at line 273 of file imagedata.cpp.

                              {
#ifndef GRAPHICS_DISABLED
  const int kTextSize = 64;
  // Draw the image.
  Pix* pix = GetPix();
  if (pix == NULL) return;
  int width = pixGetWidth(pix);
  int height = pixGetHeight(pix);
  ScrollView* win = new ScrollView("Imagedata", 100, 100,
                                   2 * (width + 2 * kTextSize),
                                   2 * (height + 4 * kTextSize),
                                   width + 10, height + 3 * kTextSize, true);
  win->Image(pix, 0, height - 1);
  pixDestroy(&pix);
  // Draw the boxes.
  win->Pen(ScrollView::RED);
  win->Brush(ScrollView::NONE);
  int text_size = kTextSize;
  if (!boxes_.empty() && boxes_[0].height() * 2 < text_size)
    text_size = boxes_[0].height() * 2;
  win->TextAttributes("Arial", text_size, false, false, false);
  if (!boxes_.empty()) {
    for (int b = 0; b < boxes_.size(); ++b) {
      boxes_[b].plot(win);
      win->Text(boxes_[b].left(), height + kTextSize, box_texts_[b].string());
    }
  } else {
    // The full transcription.
    win->Pen(ScrollView::CYAN);
    win->Text(0, height + kTextSize * 2, transcription_.string());
  }
  win->Update();
  window_wait(win);
#endif
}

GetPix()

Pix * tesseract::ImageData::GetPix

(

)

const

Definition at line 214 of file imagedata.cpp.

                             {
  return GetPixInternal(image_data_);
}

image_data()

const GenericVector<char>& tesseract::ImageData::image_data ( ) const

inline

Definition at line 137 of file imagedata.h.

                                                {
    return image_data_;
  }

imagefilename()

const STRING& tesseract::ImageData::imagefilename ( ) const

inline

Definition at line 125 of file imagedata.h.

                                      {
    return imagefilename_;
  }

language()

const STRING& tesseract::ImageData::language ( ) const

inline

Definition at line 140 of file imagedata.h.

                                 {
    return language_;
  }

MemoryUsed()

int tesseract::ImageData::MemoryUsed

(

)

const

Definition at line 268 of file imagedata.cpp.

                                {
  return image_data_.size();
}

page_number()

int tesseract::ImageData::page_number ( ) const

inline

Definition at line 131 of file imagedata.h.

                          {
    return page_number_;
  }

PreScale()

Pix * tesseract::ImageData::PreScale	(	int	target_height,
		int	max_height,
		float *	scale_factor,
		int *	scaled_width,
		int *	scaled_height,
		GenericVector< TBOX > *	boxes
	)		const

Definition at line 224 of file imagedata.cpp.

                                                           {
  int input_width = 0;
  int input_height = 0;
  Pix* src_pix = GetPix();
  ASSERT_HOST(src_pix != NULL);
  input_width = pixGetWidth(src_pix);
  input_height = pixGetHeight(src_pix);
  if (target_height == 0) {
    target_height = MIN(input_height, max_height);
  }
  float im_factor = static_cast<float>(target_height) / input_height;
  if (scaled_width != NULL)
    *scaled_width = IntCastRounded(im_factor * input_width);
  if (scaled_height != NULL)
    *scaled_height = target_height;
  // Get the scaled image.
  Pix* pix = pixScale(src_pix, im_factor, im_factor);
  if (pix == NULL) {
    tprintf("Scaling pix of size %d, %d by factor %g made null pix!!\n",
            input_width, input_height, im_factor);
  }
  if (scaled_width != NULL) *scaled_width = pixGetWidth(pix);
  if (scaled_height != NULL) *scaled_height = pixGetHeight(pix);
  pixDestroy(&src_pix);
  if (boxes != NULL) {
    // Get the boxes.
    boxes->truncate(0);
    for (int b = 0; b < boxes_.size(); ++b) {
      TBOX box = boxes_[b];
      box.scale(im_factor);
      boxes->push_back(box);
    }
    if (boxes->empty()) {
      // Make a single box for the whole image.
      TBOX box(0, 0, im_factor * input_width, target_height);
      boxes->push_back(box);
    }
  }
  if (scale_factor != NULL) *scale_factor = im_factor;
  return pix;
}

Serialize()

bool tesseract::ImageData::Serialize

(

TFile *

fp

)

const

Definition at line 165 of file imagedata.cpp.

                                         {
  if (!imagefilename_.Serialize(fp)) return false;
  if (fp->FWrite(&page_number_, sizeof(page_number_), 1) != 1) return false;
  if (!image_data_.Serialize(fp)) return false;
  if (!transcription_.Serialize(fp)) return false;
  // WARNING: Will not work across different endian machines.
  if (!boxes_.Serialize(fp)) return false;
  if (!box_texts_.SerializeClasses(fp)) return false;
  inT8 vertical = vertical_text_;
  if (fp->FWrite(&vertical, sizeof(vertical), 1) != 1) return false;
  return true;
}

set_imagefilename()

void tesseract::ImageData::set_imagefilename ( const STRING &  name )

inline

Definition at line 128 of file imagedata.h.

                                             {
    imagefilename_ = name;
  }

set_language()

void tesseract::ImageData::set_language ( const STRING &  lang )

inline

Definition at line 143 of file imagedata.h.

                                        {
    language_ = lang;
  }

set_page_number()

void tesseract::ImageData::set_page_number ( int  num )

inline

Definition at line 134 of file imagedata.h.

                                {
    page_number_ = num;
  }

SetPix()

void tesseract::ImageData::SetPix

(

Pix *

pix

)

Definition at line 209 of file imagedata.cpp.

                               {
  SetPixInternal(pix, &image_data_);
}

SkipDeSerialize()

bool tesseract::ImageData::SkipDeSerialize ( bool  swap, tesseract::TFile *  fp  )

static

Definition at line 196 of file imagedata.cpp.

                                                    {
  if (!STRING::SkipDeSerialize(swap, fp)) return false;
  inT32 page_number;
  if (fp->FRead(&page_number, sizeof(page_number), 1) != 1) return false;
  if (!GenericVector<char>::SkipDeSerialize(swap, fp)) return false;
  if (!STRING::SkipDeSerialize(swap, fp)) return false;
  if (!GenericVector<TBOX>::SkipDeSerialize(swap, fp)) return false;
  if (!GenericVector<STRING>::SkipDeSerializeClasses(swap, fp)) return false;
  inT8 vertical = 0;
  return fp->FRead(&vertical, sizeof(vertical), 1) == 1;
}

transcription()

const STRING& tesseract::ImageData::transcription ( ) const

inline

Definition at line 146 of file imagedata.h.

                                      {
    return transcription_;
  }

---

The documentation for this class was generated from the following files:

• ccstruct/imagedata.h
• ccstruct/imagedata.cpp