Smoothed histogram

Smoothed histogram, a part of algorithm transparency, is computed during feature extraction for every secondary block on the block map by summing histograms of four overlapping primary blocks from histogram data. It is used to compute clipped contrast and equalized image.



Histogram visualization can be created from this stage's binary data combined with block map. Visualization itself is not part of transparency data.

Histogram is shown separately for every secondary block on the image. Secondary blocks are smaller next to image borders. Vertical axis of histograms is logarithmic.

Binary data

Binary data structure is the same as for unsmoothed histogram except for meaning of histogram values, which now also include counts of pixels from edge-neighboring blocks.

Example: 009-smoothed-histogram.dat


Json data describes basic properties of the binary data, most importantly dimensions of the histogram. Axes x and y identify secondary block of the image. Axis bin identifies slot in the histogram for given corner.

Example: 008-smoothed-histogram.json

  "axes": [
  "dimensions": [
  "scalar": "int",
  "bitness": 32,
  "endianness": "big",
  "format": "signed"