feat: Add cpp code for two tasks

Task 1: Run tesseract and output JSON
Task 2: Fix skew on an image using leptonica

cpp/CMakeLists.txt

@@ -0,0 +1,14 @@
+cmake_minimum_required(VERSION 3.7)
+project(cpp)
+
+set(CMAKE_CXX_STANDARD 11)
+
+link_directories(/usr/local/lib/ /usr/lib/ ${Boost_LIBRARY_DIR})
+include_directories(/usr/local/include/ DocProc/src/ ${Boost_INCLUDE_DIR})
+
+set(SOURCE_FILES main.cpp main_skew.cc)
+add_executable(tessocr ${SOURCE_FILES})
+add_executable(fixskew main_skew.cc)
+
+target_link_libraries(tessocr lept tesseract opencv_core opencv_imgproc opencv_highgui opencv_imgcodecs pthread gomp docproc)
+target_link_libraries(fixskew lept tesseract opencv_core opencv_imgproc opencv_highgui opencv_imgcodecs pthread gomp docproc)

cpp/main.cpp

@@ -0,0 +1,81 @@
+#include <iostream>
+#include <docproc/binarize/binarize.h>
+#include "opencv2/core.hpp"
+#include <opencv2/imgproc.hpp>
+#include <opencv2/highgui/highgui.hpp>
+#include <memory>
+#include <docproc/utility/timer.h>
+#include <docproc/transform/transform.h>
+#include <docproc/clean/clean.h>
+#include <docproc/segment/segment.h>
+#include <docproc/post_process/post_process.h>
+#include <docproc/core/core.h>
+#include <leptonica/allheaders.h>
+#include <json/json/json.h>
+
+using namespace std;
+using namespace cv;
+
+int main(int argc, char **argv) {
+    if (argc != 3) {
+        cout << "Oops! Check arguments!" << endl;
+        return -1;
+    }
+    string imageFilePath = argv[1];
+    string jsonFilePath = argv[2];
+    cout<<jsonFilePath<<endl;
+
+    Mat originalImage = imread(imageFilePath, 1);
+
+    Mat grayScaleImage;
+    cvtColor(originalImage, grayScaleImage, CV_BGR2GRAY);
+
+    shared_ptr <tesseract::TessBaseAPI> api = shared_ptr<tesseract::TessBaseAPI>(new tesseract::TessBaseAPI());
+
+    if (api->Init(NULL, "eng", tesseract::OEM_TESSERACT_ONLY)) {
+        throw 1; // TODO: Convert to proper exception
+    }
+    api->SetVariable("tessedit_char_whitelist", "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789,.-");
+
+    api->SetImage(docproc::core::mat2PixGray(grayScaleImage)); // Gray scale is actually binary as per the UNLV dataset
+    api->Recognize(0);
+    tesseract::ResultIterator *itr = api->GetIterator();
+
+    vector<string>words;
+    vector<Rect>boundingBoxes;
+
+    {
+        tesseract::ResultIterator *ri = api->GetIterator();
+        if (ri != 0) {
+            do {
+                int left, top, right, bottom;
+                ri->BoundingBox(tesseract::RIL_WORD, &left, &top, &right, &bottom);
+                Rect rect = Rect(left, top, right - left, bottom - top);
+                words.push_back(ri->GetUTF8Text(tesseract::RIL_WORD));
+                boundingBoxes.push_back(rect);
+
+            } while (ri->Next(tesseract::RIL_WORD));
+        }
+    }
+
+    Json::Value ocrJson;
+    for (int i = 0; i < words.size(); i++) {
+        string word = words[i];
+        Rect boundingBox = boundingBoxes[i];
+        Json::Value oneWord;
+        oneWord["word"] = word;
+        Json::Value bounds;
+        bounds["x"] = boundingBox.x;
+        bounds["y"] = boundingBox.y;
+        bounds["width"] = boundingBox.width;
+        bounds["height"] = boundingBox.height;
+        oneWord["rect"] = bounds;
+
+        ocrJson[i]  = oneWord;
+    }
+    ofstream outputFileStream(jsonFilePath);
+    outputFileStream << ocrJson;
+    outputFileStream.close();
+
+    return 0;
+}

cpp/main_skew.cc

@@ -0,0 +1,33 @@
+#include <iostream>
+#include <docproc/binarize/binarize.h>
+#include "opencv2/core.hpp"
+#include <opencv2/imgproc.hpp>
+#include <opencv2/highgui/highgui.hpp>
+#include <memory>
+#include <docproc/utility/timer.h>
+#include <docproc/transform/transform.h>
+#include <docproc/clean/clean.h>
+#include <docproc/segment/segment.h>
+#include <docproc/post_process/post_process.h>
+#include <docproc/core/core.h>
+#include <leptonica/allheaders.h>
+#include <json/json/json.h>
+
+using namespace std;
+using namespace cv;
+
+int main(int argc, char **argv) {
+    if (argc != 2) {
+        cout << "Oops! Check arguments!" << endl;
+        return -1;
+    }
+    string imagePath = argv[1];
+
+    Mat binarizedImage = imread(imagePath, 0);
+
+    binarizedImage = docproc::transform::fixSkewAngle(binarizedImage);
+
+    imwrite(imagePath, binarizedImage);
+
+    return 0;
+}