Method ReadTrainingSamples

deps/tesseract/classify/mastertrainer.cpp:135–182

Reads the samples and their features from the given .tr format file, adding them to the trainer with the font_id from the content of the file. See mftraining.cpp for a description of the file format. If verification, then these are verification samples, not training.

// See mftraining.cpp for a description of the file format.
// If verification, then these are verification samples, not training.
void MasterTrainer::ReadTrainingSamples(const char* page_name,
                                        const FEATURE_DEFS_STRUCT& feature_defs,
                                        bool verification) {
  char buffer[2048];
  int int_feature_type = ShortNameToFeatureType(feature_defs, kIntFeatureType);
  int micro_feature_type = ShortNameToFeatureType(feature_defs,
                                                  kMicroFeatureType);
  int cn_feature_type = ShortNameToFeatureType(feature_defs, kCNFeatureType);
  int geo_feature_type = ShortNameToFeatureType(feature_defs, kGeoFeatureType);

  FILE* fp = Efopen(page_name, "rb");
  if (fp == NULL) {
    tprintf("Failed to open tr file: %s\n", page_name);
    return;
  }
  tr_filenames_.push_back(STRING(page_name));
  while (fgets(buffer, sizeof(buffer), fp) != NULL) {
    if (buffer[0] == '\n')
      continue;

    char* space = strchr(buffer, ' ');
    if (space == NULL) {
      tprintf("Bad format in tr file, reading fontname, unichar\n");
      continue;
    }
    *space++ = '\0';
    int font_id = GetFontInfoId(buffer);
    if (font_id < 0) font_id = 0;
    int page_number;
    STRING unichar;
    TBOX bounding_box;
    if (!ParseBoxFileStr(space, &page_number, &unichar, &bounding_box)) {
      tprintf("Bad format in tr file, reading box coords\n");
      continue;
    }
    CHAR_DESC char_desc = ReadCharDescription(feature_defs, fp);
    TrainingSample* sample = new TrainingSample;
    sample->set_font_id(font_id);
    sample->set_page_num(page_number + page_images_.size());
    sample->set_bounding_box(bounding_box);
    sample->ExtractCharDesc(int_feature_type, micro_feature_type,
                            cn_feature_type, geo_feature_type, char_desc);
    AddSample(verification, unichar.string(), sample);
    FreeCharDescription(char_desc);
  }
  charsetsize_ = unicharset_.size();
  fclose(fp);
}

// Adds the given single sample to the trainer, setting the classid
// appropriately from the given unichar_str.