Tesseract "estimating resolution as 304" 行逐字打印 + 打印时间 >1 秒。为什么以及如何压制？

Question

    tesseract::TessBaseAPI tess;
    if (tess.Init("./languages/fast/", "eng"))
    {
        std::cout << "OCRTesseract: Could not initialize tesseract." << std::endl;
        std::exit(1);
    }
    tess.SetPageSegMode(tesseract::PageSegMode::PSM_AUTO);
    tess.SetVariable("save_best_choices", "T");
    tess.SetVariable("user_defined_dpi", "70");
    auto buffer = ocrQueue.getBuffer();

    tess.SetImage((const unsigned char*)buffer->data(), buffer->getWidth(), buffer->getHeight(), 3, buffer->getWidth()*3);
    tess.Recognize(0);

    std::cout << tess.GetUTF8Text() << std::endl;

每次运行这段代码我都会得到

Estimating resolution as 304

打印时间超过 1 秒，而且一个字一个字地打印，这很烦人。

如何抑制这个？

Answer 1

tess.SetVariable("user_defined_dpi", "71");

或

tess.SetVariable("debug_file", "/dev/null");

Tesseract "estimating resolution as 304" 行逐字打印 + 打印时间 >1 秒。为什么以及如何压制？

Tesseract "estimating resolution as 304" line prints word by word + takes >1 second to print. Why and how to supress?

c++

image-processing

ocr

tesseract