vision::OCRResult missing necessary fields table_boxes and table_structure

[bingoabs]

Environment

FastDeploy version: 1.0.7
OS Platform: Windows x64 intel
Hardware: Nvidia GPU 3080Ti CUDA 11.2 CUDNN 8.3
Program Language: . Python 3.10

Problem description

The vision::OCRResult actually include table_boxes and table_structure fields, but the pybing11 part doesn't include them. We can find the both fields in the first code snippet below, but not in the second one.

// FastDeploy\fastdeploy\vision\common\result.h
struct FASTDEPLOY_DECL OCRResult : public BaseResult {
  std::vector<std::array<int, 8>> boxes;

  std::vector<std::string> text;
  std::vector<float> rec_scores;

  std::vector<float> cls_scores;
  std::vector<int32_t> cls_labels;

  std::vector<std::array<int, 8>> table_boxes;
  std::vector<std::string> table_structure;
  std::string table_html;

  ResultType type = ResultType::OCR;

  void Clear();

  std::string Str();
};

  pybind11::class_<vision::OCRResult>(m, "OCRResult")
      .def(pybind11::init())
      .def_readwrite("boxes", &vision::OCRResult::boxes)
      .def_readwrite("text", &vision::OCRResult::text)
      .def_readwrite("rec_scores", &vision::OCRResult::rec_scores)
      .def_readwrite("cls_scores", &vision::OCRResult::cls_scores)
      .def_readwrite("cls_labels", &vision::OCRResult::cls_labels)
      .def("__repr__", &vision::OCRResult::Str)
      .def("__str__", &vision::OCRResult::Str);

Solution

• The directly way to solve it is to add the two fields to pybind11 part. Or we can derive a TableOCRResult class. Thanks.

[juncaipeng]

Thank you for raising this issue. We will consider fixing it in the future