report op• Data kinds: metrics → table
• Call: import fullseye as fs; fs.ledger.metrics_table(report, order=None) (to call the implementation directly, import imgmetrics; imgmetrics.metrics_table(report, order=None); from the registry, opsimgmetrics.get("metrics_table"))
• GPU: none (CPU reference implementation only)
Turn a report into a (name, value) table — the rows describing the conditions are always listed with it.
The gist of this op is to make it impossible to create tables with only numerical values. contract Each item is in the same table as Condition: <name>.
• report: return value of compare_images or measure_with (contract dict) with keys. otherwise MetricContractError.
• order: Order of index names to display. When specified, indexes that are not here are dropped, and when omitted, the keys are in dictionary order. Not in report The name is silently skipped, and "contract" is not listed as an indicator.
• Returns: a list of (name, value) tuples. After the index line, each item in contract Add keys in lexicographical order like "condition: data_range". do not convert the value (None and inf are also the same).
Place it just before adding it to the figure annotation or Markdown table. Do not paste after deleting the condition line -- When two PSNRs with different data_range are placed side by side, there is an apparent difference of 48 dB.
• image_difference_metrics family guide
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• image_quality_metrics — py -3.11 examples/image_quality_metrics.py
table as input)luma_limited_u8 · rank_data · tid2013_index · tid2013_by_distortion
report)compare_images · measure_with · data_range_of
*Provenance: imgmetrics.py — IMGMETRICS operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.