cntraining.cpp File Reference

#include "oldlist.h"
#include "efio.h"
#include "emalloc.h"
#include "featdefs.h"
#include "tessopt.h"
#include "ocrfeatures.h"
#include "clusttool.h"
#include "cluster.h"
#include <string.h>
#include <stdio.h>
#include <math.h>
#include "unichar.h"
#include "commontraining.h"

Go to the source code of this file.

Macros
#define	PROGRAM_FEATURE_TYPE   "cn"

Functions
	DECLARE_STRING_PARAM_FLAG (D)
int	main (int argc, char **argv)
void	WriteNormProtos (const char *Directory, LIST LabeledProtoList, CLUSTERER *Clusterer)
void	WriteProtos (FILE *File, uinT16 N, LIST ProtoList, BOOL8 WriteSigProtos, BOOL8 WriteInsigProtos)
int	main (int argc, char *argv[])

Variables
CLUSTERCONFIG	CNConfig

Macro Definition Documentation

#define PROGRAM_FEATURE_TYPE   "cn"

Definition at line 41 of file cntraining.cpp.

Function Documentation

DECLARE_STRING_PARAM_FLAG

(

D

)

int main	(	int	argc,
		char **	argv
	)

This program reads in a text file consisting of feature samples from a training page in the following format:

   FontName UTF8-char-str xmin ymin xmax ymax page-number
    NumberOfFeatureTypes(N)
      FeatureTypeName1 NumberOfFeatures(M)
         Feature1
         ...
         FeatureM
      FeatureTypeName2 NumberOfFeatures(M)
         Feature1
         ...
         FeatureM
      ...
      FeatureTypeNameN NumberOfFeatures(M)
         Feature1
         ...
         FeatureM
   FontName CharName ...

The result of this program is a binary inttemp file used by the OCR engine.

Parameters

argc	number of command line arguments
argv	array of command line arguments

Returns

none

Note

Exceptions: none

History: Fri Aug 18 08:56:17 1989, DSJ, Created.

History: Mon May 18 1998, Christy Russson, Revistion started.

Definition at line 41 of file tesseractmain.cpp.

                                {
  if ((argc == 2 && strcmp(argv[1], "-v") == 0) ||
      (argc == 2 && strcmp(argv[1], "--version") == 0)) {
    char *versionStrP;

    fprintf(stderr, "tesseract %s\n", tesseract::TessBaseAPI::Version());

    versionStrP = getLeptonicaVersion();
    fprintf(stderr, " %s\n", versionStrP);
    lept_free(versionStrP);

    versionStrP = getImagelibVersions();
    fprintf(stderr, "  %s\n", versionStrP);
    lept_free(versionStrP);

#ifdef USE_OPENCL
    cl_platform_id platform;
    cl_uint num_platforms;
    cl_device_id devices[2];
    cl_uint num_devices;
    char info[256];
    int i;

    fprintf(stderr, " OpenCL info:\n");
    clGetPlatformIDs(1, &platform, &num_platforms);
    fprintf(stderr, "  Found %d platforms.\n", num_platforms);
    clGetPlatformInfo(platform, CL_PLATFORM_NAME, 256, info, 0);
    fprintf(stderr, "  Platform name: %s.\n", info);
    clGetPlatformInfo(platform, CL_PLATFORM_VERSION, 256, info, 0);
    fprintf(stderr, "  Version: %s.\n", info);
    clGetDeviceIDs(platform, CL_DEVICE_TYPE_ALL, 2, devices, &num_devices);
    fprintf(stderr, "  Found %d devices.\n", num_devices);
    for (i = 0; i < num_devices; ++i) {
      clGetDeviceInfo(devices[i], CL_DEVICE_NAME, 256, info, 0);
      fprintf(stderr, "    Device %d name: %s.\n", i+1, info);
    }
#endif
    exit(0);
  }

  // Make the order of args a bit more forgiving than it used to be.
  const char* lang = "eng";
  const char* image = NULL;
  const char* outputbase = NULL;
  const char* datapath = NULL;
  bool noocr = false;
  bool list_langs = false;
  bool print_parameters = false;
  GenericVector<STRING> vars_vec, vars_values;

  tesseract::PageSegMode pagesegmode = tesseract::PSM_AUTO;
  int arg = 1;
  while (arg < argc && (outputbase == NULL || argv[arg][0] == '-')) {
    if (strcmp(argv[arg], "-l") == 0 && arg + 1 < argc) {
      lang = argv[arg + 1];
      ++arg;
    } else if (strcmp(argv[arg], "--tessdata-dir") == 0 && arg + 1 < argc) {
      datapath = argv[arg + 1];
      ++arg;
    } else if (strcmp(argv[arg], "--user-words") == 0 && arg + 1 < argc) {
      vars_vec.push_back("user_words_file");
      vars_values.push_back(argv[arg + 1]);
      ++arg;
    } else if (strcmp(argv[arg], "--user-patterns") == 0 && arg + 1 < argc) {
      vars_vec.push_back("user_patterns_file");
      vars_values.push_back(argv[arg + 1]);
      ++arg;
    } else if (strcmp(argv[arg], "--list-langs") == 0) {
      noocr = true;
      list_langs = true;
    } else if (strcmp(argv[arg], "-psm") == 0 && arg + 1 < argc) {
      pagesegmode = static_cast<tesseract::PageSegMode>(atoi(argv[arg + 1]));
      ++arg;
    } else if (strcmp(argv[arg], "--print-parameters") == 0) {
      noocr = true;
      print_parameters = true;
    } else if (strcmp(argv[arg], "-c") == 0 && arg + 1 < argc) {
      // handled properly after api init
      ++arg;
    } else if (image == NULL) {
      image = argv[arg];
    } else if (outputbase == NULL) {
      outputbase = argv[arg];
    }
    ++arg;
  }

  if (argc == 2 && strcmp(argv[1], "--list-langs") == 0) {
    list_langs = true;
    noocr = true;
  }

  if (outputbase == NULL && noocr == false) {
    fprintf(stderr, "Usage:\n  %s imagename|stdin outputbase|stdout "
            "[options...] [configfile...]\n\n", argv[0]);

    fprintf(stderr, "OCR options:\n");
    fprintf(stderr, "  --tessdata-dir /path\tspecify the location of tessdata"
                      " path\n");
    fprintf(stderr, "  --user-words /path/to/file\tspecify the location of user"
            " words file\n");
    fprintf(stderr, "  --user-patterns /path/to/file\tspecify the location of"
            " user patterns file\n");
    fprintf(stderr, "  -l lang[+lang]\tspecify language(s) used for OCR\n");
    fprintf(stderr, "  -c configvar=value\tset value for control parameter.\n"
                      "\t\t\tMultiple -c arguments are allowed.\n");
    fprintf(stderr, "  -psm pagesegmode\tspecify page segmentation mode.\n");
    fprintf(stderr, "These options must occur before any configfile.\n\n");
    fprintf(stderr,
            "pagesegmode values are:\n"
            "  0 = Orientation and script detection (OSD) only.\n"
            "  1 = Automatic page segmentation with OSD.\n"
            "  2 = Automatic page segmentation, but no OSD, or OCR\n"
            "  3 = Fully automatic page segmentation, but no OSD. (Default)\n"
            "  4 = Assume a single column of text of variable sizes.\n"
            "  5 = Assume a single uniform block of vertically aligned text.\n"
            "  6 = Assume a single uniform block of text.\n"
            "  7 = Treat the image as a single text line.\n"
            "  8 = Treat the image as a single word.\n"
            "  9 = Treat the image as a single word in a circle.\n"
            "  10 = Treat the image as a single character.\n\n");
    fprintf(stderr, "Single options:\n");
    fprintf(stderr, "  -v --version: version info\n");
    fprintf(stderr, "  --list-langs: list available languages for tesseract "
                      "engine. Can be used with --tessdata-dir.\n");
    fprintf(stderr, "  --print-parameters: print tesseract parameters to the "
                      "stdout.\n");
    exit(1);
  }

  if (outputbase != NULL && strcmp(outputbase, "-") &&
      strcmp(outputbase, "stdout")) {
    tprintf("Tesseract Open Source OCR Engine v%s with Leptonica\n",
           tesseract::TessBaseAPI::Version());
  }
  PERF_COUNT_START("Tesseract:main")
  tesseract::TessBaseAPI api;

  api.SetOutputName(outputbase);
  int rc = api.Init(datapath, lang, tesseract::OEM_DEFAULT,
                &(argv[arg]), argc - arg, &vars_vec, &vars_values, false);

  if (rc) {
    fprintf(stderr, "Could not initialize tesseract.\n");
    exit(1);
  }

  char opt1[255], opt2[255];
  for (arg = 0; arg < argc; arg++) {
    if (strcmp(argv[arg], "-c") == 0 && arg + 1 < argc) {
      strncpy(opt1, argv[arg + 1], 255);
      char *p = strchr(opt1, '=');
      if (!p) {
        fprintf(stderr, "Missing = in configvar assignment\n");
        exit(1);
      }
      *p = 0;
      strncpy(opt2, strchr(argv[arg + 1], '=') + 1, 255);
      opt2[254] = 0;
      ++arg;

      if (!api.SetVariable(opt1, opt2)) {
        fprintf(stderr, "Could not set option: %s=%s\n", opt1, opt2);
      }
    }
  }

  if (list_langs) {
     GenericVector<STRING> languages;
     api.GetAvailableLanguagesAsVector(&languages);
     fprintf(stderr, "List of available languages (%d):\n",
             languages.size());
     for (int index = 0; index < languages.size(); ++index) {
       STRING& string = languages[index];
       fprintf(stderr, "%s\n", string.string());
     }
     api.End();
     exit(0);
  }

  if (print_parameters) {
     FILE* fout = stdout;
     fprintf(stdout, "Tesseract parameters:\n");
     api.PrintVariables(fout);
     api.End();
     exit(0);
  }

  // We have 2 possible sources of pagesegmode: a config file and
  // the command line. For backwards compatability reasons, the
  // default in tesseract is tesseract::PSM_SINGLE_BLOCK, but the
  // default for this program is tesseract::PSM_AUTO. We will let
  // the config file take priority, so the command-line default
  // can take priority over the tesseract default, so we use the
  // value from the command line only if the retrieved mode
  // is still tesseract::PSM_SINGLE_BLOCK, indicating no change
  // in any config file. Therefore the only way to force
  // tesseract::PSM_SINGLE_BLOCK is from the command line.
  // It would be simpler if we could set the value before Init,
  // but that doesn't work.
  if (api.GetPageSegMode() == tesseract::PSM_SINGLE_BLOCK)
     api.SetPageSegMode(pagesegmode);

  if (pagesegmode == tesseract::PSM_AUTO_ONLY ||
      pagesegmode == tesseract::PSM_OSD_ONLY) {
    int ret_val = 0;

    Pix* pixs = pixRead(image);
    if (!pixs) {
      fprintf(stderr, "Cannot open input file: %s\n", image);
      exit(2);
    }
    api.SetImage(pixs);

    if (pagesegmode == tesseract::PSM_OSD_ONLY) {
       OSResults osr;
       if (api.DetectOS(&osr)) {
         int orient = osr.best_result.orientation_id;
         int script_id = osr.get_best_script(orient);
         float orient_oco = osr.best_result.oconfidence;
         float orient_sco = osr.best_result.sconfidence;
         tprintf("Orientation: %d\nOrientation in degrees: %d\n" \
                 "Orientation confidence: %.2f\n" \
                 "Script: %d\nScript confidence: %.2f\n",
                 orient, OrientationIdToValue(orient), orient_oco,
                 script_id, orient_sco);
       } else {
         ret_val = 1;
       }
    } else {
       tesseract::Orientation orientation;
       tesseract::WritingDirection direction;
       tesseract::TextlineOrder order;
       float deskew_angle;
       tesseract::PageIterator* it =  api.AnalyseLayout();
       if (it) {
         it->Orientation(&orientation, &direction, &order, &deskew_angle);
         tprintf("Orientation: %d\nWritingDirection: %d\nTextlineOrder: %d\n" \
                 "Deskew angle: %.4f\n",
                 orientation, direction, order, deskew_angle);
       } else {
         ret_val = 1;
       }
       delete it;
    }
    pixDestroy(&pixs);
    exit(ret_val);
  }

  bool b;
  tesseract::PointerVector<tesseract::TessResultRenderer> renderers;
  api.GetBoolVariable("tessedit_create_hocr", &b);
  if (b) {
    bool font_info;
    api.GetBoolVariable("hocr_font_info", &font_info);
    renderers.push_back(new tesseract::TessHOcrRenderer(outputbase, font_info));
  }
  api.GetBoolVariable("tessedit_create_pdf", &b);
  if (b) {
    renderers.push_back(new tesseract::TessPDFRenderer(outputbase,
                                                       api.GetDatapath()));
  }
  api.GetBoolVariable("tessedit_write_unlv", &b);
  if (b) renderers.push_back(new tesseract::TessUnlvRenderer(outputbase));
  api.GetBoolVariable("tessedit_create_boxfile", &b);
  if (b) renderers.push_back(new tesseract::TessBoxTextRenderer(outputbase));
  api.GetBoolVariable("tessedit_create_txt", &b);
  if (b) renderers.push_back(new tesseract::TessTextRenderer(outputbase));
  if (!renderers.empty()) {
    // Since the PointerVector auto-deletes, null-out the renderers that are
    // added to the root, and leave the root in the vector.
    for (int r = 1; r < renderers.size(); ++r) {
      renderers[0]->insert(renderers[r]);
      renderers[r] = NULL;
    }
    if (!api.ProcessPages(image, NULL, 0, renderers[0])) {
      fprintf(stderr, "Error during processing.\n");
      exit(1);
    }
  }

  PERF_COUNT_END
  return 0;                      // Normal exit
}

int main	(	int	argc,
		char *	argv[]
	)

This program reads in a text file consisting of feature samples from a training page in the following format:

   FontName CharName NumberOfFeatureTypes(N)
      FeatureTypeName1 NumberOfFeatures(M)
         Feature1
         ...
         FeatureM
      FeatureTypeName2 NumberOfFeatures(M)
         Feature1
         ...
         FeatureM
      ...
      FeatureTypeNameN NumberOfFeatures(M)
         Feature1
         ...
         FeatureM
   FontName CharName ...

It then appends these samples into a separate file for each character. The name of the file is

DirectoryName/FontName/CharName.FeatureTypeName

The DirectoryName can be specified via a command line argument. If not specified, it defaults to the current directory. The format of the resulting files is:

   NumberOfFeatures(M)
      Feature1
      ...
      FeatureM
   NumberOfFeatures(M)
   ...

The output files each have a header which describes the type of feature which the file contains. This header is in the format required by the clusterer. A command line argument can also be used to specify that only the first N samples of each class should be used.

Parameters

argc	number of command line arguments
argv	array of command line arguments

Returns

none

Note

Globals: none

Exceptions: none

History: Fri Aug 18 08:56:17 1989, DSJ, Created.

Definition at line 137 of file cntraining.cpp.

{
  // Set the global Config parameters before parsing the command line.
  Config = CNConfig;

  const char  *PageName;
  FILE  *TrainingPage;
  LIST  CharList = NIL_LIST;
  CLUSTERER  *Clusterer = NULL;
  LIST    ProtoList = NIL_LIST;
  LIST    NormProtoList = NIL_LIST;
  LIST pCharList;
  LABELEDLIST CharSample;
  FEATURE_DEFS_STRUCT FeatureDefs;
  InitFeatureDefs(&FeatureDefs);

  ParseArguments(&argc, &argv);
  int num_fonts = 0;
  while ((PageName = GetNextFilename(argc, argv)) != NULL) {
    printf("Reading %s ...\n", PageName);
    TrainingPage = Efopen(PageName, "rb");
    ReadTrainingSamples(FeatureDefs, PROGRAM_FEATURE_TYPE,
                        100, NULL, TrainingPage, &CharList);
    fclose(TrainingPage);
    ++num_fonts;
  }
  printf("Clustering ...\n");
  // To allow an individual font to form a separate cluster,
  // reduce the min samples:
  // Config.MinSamples = 0.5 / num_fonts;
  pCharList = CharList;
  iterate(pCharList) {
    //Cluster
    CharSample = (LABELEDLIST)first_node(pCharList);
    Clusterer =
      SetUpForClustering(FeatureDefs, CharSample, PROGRAM_FEATURE_TYPE);
    float SavedMinSamples = Config.MinSamples;
    // To disable the tendency to produce a single cluster for all fonts,
    // make MagicSamples an impossible to achieve number:
    // Config.MagicSamples = CharSample->SampleCount * 10;
    Config.MagicSamples = CharSample->SampleCount;
    while (Config.MinSamples > 0.001) {
      ProtoList = ClusterSamples(Clusterer, &Config);
      if (NumberOfProtos(ProtoList, 1, 0) > 0) {
        break;
      } else {
        Config.MinSamples *= 0.95;
        printf("0 significant protos for %s."
               " Retrying clustering with MinSamples = %f%%\n",
               CharSample->Label, Config.MinSamples);
      }
    }
    Config.MinSamples = SavedMinSamples;
    AddToNormProtosList(&NormProtoList, ProtoList, CharSample->Label);
  }
  FreeTrainingSamples(CharList);
  if (Clusterer == NULL) { // To avoid a SIGSEGV
    fprintf(stderr, "Error: NULL clusterer!\n");
    return 1;
  }
  WriteNormProtos(FLAGS_D.c_str(), NormProtoList, Clusterer);
  FreeNormProtoList(NormProtoList);
  FreeProtoList(&ProtoList);
  FreeClusterer(Clusterer);
  printf ("\n");
  return 0;
}  // main

void WriteNormProtos	(	const char *	Directory,
		LIST	LabeledProtoList,
		CLUSTERER *	Clusterer
	)

This routine writes the specified samples into files which are organized according to the font name and character name of the samples.

Parameters

Directory	directory to place sample files into
LabeledProtoList	List of labeled protos
Clusterer	The CLUSTERER to use

Returns

none

Note

Exceptions: none

History: Fri Aug 18 16:17:06 1989, DSJ, Created.

Definition at line 222 of file cntraining.cpp.

{
  FILE    *File;
  STRING Filename;
  LABELEDLIST LabeledProto;
  int N;

  Filename = "";
  if (Directory != NULL && Directory[0] != '\0')
  {
    Filename += Directory;
    Filename += "/";
  }
  Filename += "normproto";
  printf ("\nWriting %s ...", Filename.string());
  File = Efopen (Filename.string(), "wb");
  fprintf(File,"%0d\n",Clusterer->SampleSize);
  WriteParamDesc(File,Clusterer->SampleSize,Clusterer->ParamDesc);
  iterate(LabeledProtoList)
  {
    LabeledProto = (LABELEDLIST) first_node (LabeledProtoList);
    N = NumberOfProtos(LabeledProto->List, true, false);
    if (N < 1) {
      printf ("\nError! Not enough protos for %s: %d protos"
              " (%d significant protos"
              ", %d insignificant protos)\n",
              LabeledProto->Label, N,
              NumberOfProtos(LabeledProto->List, 1, 0),
              NumberOfProtos(LabeledProto->List, 0, 1));
      exit(1);
    }
    fprintf(File, "\n%s %d\n", LabeledProto->Label, N);
    WriteProtos(File, Clusterer->SampleSize, LabeledProto->List, true, false);
  }
  fclose (File);

}  // WriteNormProtos

void WriteProtos	(	FILE *	File,
		uinT16	N,
		LIST	ProtoList,
		BOOL8	WriteSigProtos,
		BOOL8	WriteInsigProtos
	)

Definition at line 264 of file cntraining.cpp.

{
  PROTOTYPE  *Proto;

  // write prototypes
  iterate(ProtoList)
  {
    Proto = (PROTOTYPE *) first_node ( ProtoList );
    if (( Proto->Significant && WriteSigProtos )  ||
      ( ! Proto->Significant && WriteInsigProtos ) )
      WritePrototype( File, N, Proto );
  }
}  // WriteProtos

Variable Documentation

CLUSTERCONFIG CNConfig

Initial value:

=
{
  elliptical, 0.025, 0.05, 0.8, 1e-3, 0
}

Definition at line 79 of file cntraining.cpp.