Noddy little Tesseract wrapper.

tesseractcli.cpp

// ----------------------------------------------------------------------------
//  Noddy little Tesseract wrapper.
//
//  Usage:
//  
//      ./tesseractcli <path to screenshot image> [left] [top] [width] [height]
//      
//      -   Screenshot path is mandatory, can be relative to current working
//          directory or absolute.
//      -   Left, top, width, and height are optional integers that set a
//          bounding rectable on the screenshot within which to detect text.
//      
//  Returns:
//  
//      -   Exit code 0 with detected text to stdout if OK.
//      -   Else exit code 1 with error message to stderr.
//      
//  How to compile:
//  
//      yum groupinstall 'Development Tools'
//      yum install tesseract tesseract-devel leptonica leptonica-devel
//      g++ -pthread $(pkg-config --cflags --libs tesseract lept) \
//          tesseractcli.cpp -o tesseractcli
//          
//  Notes:
//      -   The default tesseract CLI tool only writes output to a filepath,
//          which is unnecessary, so this CLI tool outputs the text to stdout.
//      -   The default tesseract CLI tool doesn't let you set a rectangle
//          within which to detect text. This one does.
// ----------------------------------------------------------------------------

#include <iostream>
#include <string>
#include <sys/stat.h>
#include <sys/types.h>
#include <unistd.h>

#include <leptonica/allheaders.h>
#include <tesseract/baseapi.h>

bool does_file_exist(const char *path) {
    struct stat buffer;
    return stat(path, &buffer) == 0;
}

int main(int argc, char *argv[]) {
    if (argc == 1) {
        std::cerr << "Need path to screenshot image." << std::endl;
        exit(1);
    }
    if (!(does_file_exist(argv[1]))) {
        std::cerr << "Path '" << argv[1] << "' does not exist" << std::endl;
        exit(1);        
    }
    tesseract::TessBaseAPI *api = new tesseract::TessBaseAPI();
    // Initialize tesseract-ocr with English, without specifying tessdata path
    if (api->Init(NULL, "eng")) {
        std::cerr << "Could not initialize tesseract." << std::endl;
        exit(1);
    }

    // Open input image with leptonica library
    Pix *image = pixRead(argv[1]);
    api->SetImage(image);    
    if (argc > 2) {
        int left = atoi(argv[2]), 
            top = atoi(argv[3]),
            width = atoi(argv[4]),
            height = atoi(argv[5]);
        api->SetRectangle(left, top, width, height);
    }

    // Get OCR result
    char *outText = api->GetUTF8Text();
    std::cout << outText << std::endl;

    // Destroy used object and release memory
    api->End();
    delete [] outText;
    pixDestroy(&image);

    return 0;
}