1 // Copyright 2011 Google Inc. All Rights Reserved.
2 // Author: rays@google.com (Ray Smith)
3
4 // Licensed under the Apache License, Version 2.0 (the "License");
5 // you may not use this file except in compliance with the License.
6 // You may obtain a copy of the License at
7 // http://www.apache.org/licenses/LICENSE-2.0
8 // Unless required by applicable law or agreed to in writing, software
9 // distributed under the License is distributed on an "AS IS" BASIS,
10 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
11 // See the License for the specific language governing permissions and
12 // limitations under the License.
13
14 // Filename: classifier_tester.cpp
15 // Purpose: Tests a character classifier on data as formatted for training,
16 // but doesn't have to be the same as the training data.
17 // Author: Ray Smith
18
19 #include <tesseract/baseapi.h>
20 #include <algorithm>
21 #include <cstdio>
22 #include "commontraining.h"
23 #include "mastertrainer.h"
24 #include "params.h"
25 #include "tessclassifier.h"
26 #include "tesseractclass.h"
27
28 using namespace tesseract;
29
30 static STRING_PARAM_FLAG(classifier, "", "Classifier to test");
31 static STRING_PARAM_FLAG(lang, "eng", "Language to test");
32 static STRING_PARAM_FLAG(tessdata_dir, "", "Directory of traineddata files");
33
34 enum ClassifierName { CN_PRUNER, CN_FULL, CN_COUNT };
35
36 static const char *names[] = {"pruner", "full"};
37
InitializeClassifier(const char * classifer_name,const UNICHARSET & unicharset,int argc,char ** argv,tesseract::TessBaseAPI ** api)38 static tesseract::ShapeClassifier *InitializeClassifier(const char *classifer_name,
39 const UNICHARSET &unicharset, int argc,
40 char **argv, tesseract::TessBaseAPI **api) {
41 // Decode the classifier string.
42 ClassifierName classifier = CN_COUNT;
43 for (int c = 0; c < CN_COUNT; ++c) {
44 if (strcmp(classifer_name, names[c]) == 0) {
45 classifier = static_cast<ClassifierName>(c);
46 break;
47 }
48 }
49 if (classifier == CN_COUNT) {
50 fprintf(stderr, "Invalid classifier name:%s\n", FLAGS_classifier.c_str());
51 return nullptr;
52 }
53
54 // We need to initialize tesseract to test.
55 *api = new tesseract::TessBaseAPI;
56 tesseract::OcrEngineMode engine_mode = tesseract::OEM_TESSERACT_ONLY;
57 tesseract::Tesseract *tesseract = nullptr;
58 tesseract::Classify *classify = nullptr;
59 if (classifier == CN_PRUNER || classifier == CN_FULL) {
60 if ((*api)->Init(FLAGS_tessdata_dir.c_str(), FLAGS_lang.c_str(), engine_mode) < 0) {
61 fprintf(stderr, "Tesseract initialization failed!\n");
62 return nullptr;
63 }
64 tesseract = const_cast<tesseract::Tesseract *>((*api)->tesseract());
65 classify = static_cast<tesseract::Classify *>(tesseract);
66 if (classify->shape_table() == nullptr) {
67 fprintf(stderr, "Tesseract must contain a ShapeTable!\n");
68 return nullptr;
69 }
70 }
71 tesseract::ShapeClassifier *shape_classifier = nullptr;
72
73 if (classifier == CN_PRUNER) {
74 shape_classifier = new tesseract::TessClassifier(true, classify);
75 } else if (classifier == CN_FULL) {
76 shape_classifier = new tesseract::TessClassifier(false, classify);
77 }
78 tprintf("Testing classifier %s:\n", classifer_name);
79 return shape_classifier;
80 }
81
82 // This program has complex setup requirements, so here is some help:
83 // Two different modes, tr files and serialized mastertrainer.
84 // From tr files:
85 // classifier_tester -U unicharset -F font_properties -X xheights
86 // -classifier x -lang lang [-output_trainer trainer] *.tr
87 // From a serialized trainer:
88 // classifier_tester -input_trainer trainer [-lang lang] -classifier x
89 //
90 // In the first case, the unicharset must be the unicharset from within
91 // the classifier under test, and the font_properties and xheights files must
92 // match the files used during training.
93 // In the second case, the trainer file must have been prepared from
94 // some previous run of shapeclustering, mftraining, or classifier_tester
95 // using the same conditions as above, ie matching unicharset/font_properties.
96 //
97 // Available values of classifier (x above) are:
98 // pruner : Tesseract class pruner only.
99 // full : Tesseract full classifier.
100 // with an input trainer.)
main(int argc,char ** argv)101 int main(int argc, char **argv) {
102 tesseract::CheckSharedLibraryVersion();
103 ParseArguments(&argc, &argv);
104 std::string file_prefix;
105 auto trainer = tesseract::LoadTrainingData(argv + 1, false, nullptr, file_prefix);
106 tesseract::TessBaseAPI *api;
107 // Decode the classifier string.
108 tesseract::ShapeClassifier *shape_classifier =
109 InitializeClassifier(FLAGS_classifier.c_str(), trainer->unicharset(), argc, argv, &api);
110 if (shape_classifier == nullptr) {
111 fprintf(stderr, "Classifier init failed!:%s\n", FLAGS_classifier.c_str());
112 return 1;
113 }
114
115 // We want to test junk as well if it is available.
116 // trainer->IncludeJunk();
117 // We want to test with replicated samples too.
118 trainer->ReplicateAndRandomizeSamplesIfRequired();
119
120 trainer->TestClassifierOnSamples(tesseract::CT_UNICHAR_TOP1_ERR,
121 std::max(3, static_cast<int>(FLAGS_debug_level)), false,
122 shape_classifier, nullptr);
123 delete shape_classifier;
124 delete api;
125
126 return 0;
127 } /* main */
128