CUDA implementation wants to convert std::vector<KeyPoint> <-> GpuMat.
There is no direct mapping from KeyPoint (mix of int/float fields)
into cv::Mat element type, so this conversion must be avoided.

Legacy mode is turned back for CUDA builds.

DMatch and Keypoint are not compatible types (mixed float/int fields)

Setting truncate option for QString leftJustified method to false fixes 10 character truncation issue noted in issue 9489

size_t is not supported cv::Mat type

  * guard correctly in header file
  * guard correctly in cmake file