Block orientation (SourceAFIS transparency data)

SourceAFIS » Algorithm » Transparency » Block orientation

Block orientation, a part of algorithm transparency, contains orientation vector for primary blocks on the block map that are within mask. Orientation vector has the same meaning as in pixelwise orientation, from which block orientation is computed during feature extraction by adding up orientation vectors within every block. Block orientation is refined into smoothed orientation.



This visualization was constructed from this stage's binary data combined with block map, filtered mask, and the original fingerprint image. Visualization itself is not part of transparency data.

Block-level ridge orientation.

Binary data

Structure of binary data is identical to that of pixelwise orientation except that orientation is tracked per primary block rather than per pixel. No orientation is computed for blocks outside the mask.

Example: 026-block-orientation.dat (1,300 scalars, range -472 to 641, average -48.1)


Json data describes basic properties of the binary data, most importantly dimensions of the orientation map, which is the same as the number of primary blocks.

Example: 025-block-orientation.json

  "axes": [
  "dimensions": [
  "scalar": "double",
  "bitness": 64,
  "endianness": "big",
  "format": "IEEE754"