std::vector<std::thread> workers;
    unsigned threads = std::thread::hardware_concurrency();
    unsigned line_start = 0;
    unsigned line_adv = _size.hei / threads;
    for(unsigned i=0; i<threads; ++i) {
        workers.emplace_back(processBlock, line_start, line_start + line_adv);
        line_start += line_adv;
    }

    std::for_each(workers.begin(), workers.end(), [](std::thread &t)  {
        t.join();
    });