package sparse_mapping_protobuf

Get desktop application:
View/edit binary Protocol Buffers messages

Used in: Frame

• required double r00 = 1

• required double r01 = 2

• required double r02 = 3

• required double r10 = 4

• required double r11 = 5

• required double r12 = 6

• required double r20 = 7

• required double r21 = 8

• required double r22 = 9

• required double t0 = 10

• required double t1 = 11

• required double t2 = 12

Used in: Map

• repeated double focal_length = 1

  Focal length is a vector x, y. In pixels. This length is the same for both distorted and undistorted imagery.

• repeated double optical_offset = 2

  Optical offset is the location of the optical axis in the distorted coordinate frame or 'DISTORTED' in camera params lingo.

• repeated int32 distorted_image_size = 3

  Image size in pixels

• repeated int32 undistorted_image_size = 4

  Image size in pixels. This will be the same as in distorted_image_size if the camera model was already undistorted or 'metric'.

• repeated double distortion = 5

  Distortion vector comes in multiple sizes. Size 0, there is no lens distortion. Size 1, this is an FOV/Tangent lens distortion. Size 4 and 5 are TSAI OpenCV coefficients.

Used in: Landmark

• required double r00 = 1

• required double r01 = 2

• required double r02 = 3

• required double r11 = 4

• required double r12 = 5

• required double r22 = 6

• required int32 num_entries = 1

• required int32 num_inverted_index = 4

• required int32 word_id = 1

• required int32 entry_id = 2

• required double weight = 3

• required int32 node_id = 1

• required int32 parent_id = 2

• required double weight = 3

• required bytes feature = 4

these fields are all defined in TemplatedVocabulary and TempaltedDatabase from DBoW

• required int32 k = 1

• required int32 l = 2

• required int32 scoring_type = 3

  enum

• required int32 weighting_type = 4

  enum

• required int32 num_nodes = 5

• required int32 num_words = 6

• required int32 word_id = 1

• required int32 node_id = 2

Used in: Frame

• required double x = 1

• required double y = 2

• required bytes description = 3

• optional float scale = 4

• optional float orientation = 5

• optional string name = 1

• repeated Feature feature = 2

• optional Affine3d pose = 3

• required Vector3d loc = 1

• repeated Matching match = 2

• optional Covariance cov = 3

• optional string descriptor_name = 1

  this is no longer used, only detector_name is needed

• required string detector_name = 2

• required CameraModel camera = 3

• required int32 descriptor_depth = 4

  This is the same as cv::Mat::depth()

• required int32 num_frames = 5

• required int32 num_landmarks = 6

• optional Map.VocabDB vocab_db = 7

• optional int32 orgbrisk_threshold = 8

  this is no longer used but remains for compatability

• optional int32 histogram_equalization = 9

  histogram_equalization 1 means true, 0 means false, 2 means not known

Used in: Map

• NONE = 0

• BINARYDB = 1

• FLOATDB = 2

• VOCABTREE = 3

Used in: Landmark

• required int32 camera_id = 1

• required int32 feature_id = 2

Used in: Landmark

• required double x = 1

• required double y = 2

• required double z = 3