- base_connaissances/ → snippets/ (corpus/ → textes/, instructions.md → consignes_redaction.md, decoupage.py → fichiers.py, importer.py → envoyer.py) ; python3 -m snippets.envoyer. - instance-de-test/ : 1_demarrer.sh, 2_creer_boite_et_cle.py, 3_brancher_ia.py, 4_essayer_brouillons.py, voir_empreintes.sh, connexion_admin.py. - ollama/verifier.py → ollama/verifier_ollama.py ; tests, tâches VS Code et docs alignés. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
14 lines
862 B
Markdown
14 lines
862 B
Markdown
# Station Guide — Target Image Quality and Pattern-Matching Metrics
|
|
|
|
These metrics are calculated independently for each processed target image. They describe
|
|
how clearly the target is visible, how stable its image appearance remains, and how reliably the
|
|
current target image can be matched with its reference pattern.
|
|
They are primarily diagnostic indicators intended to explain changes in measurement quality and
|
|
help identify their likely cause. Most metrics should be interpreted as trends over time and in
|
|
combination with the corresponding image and other available indicators, rather than as
|
|
standalone pass/fail criteria.
|
|
The Error Value is a specific exception: when an Error Threshold is configured, it is also used to
|
|
identify and reject measurements for which pattern matching is considered unreliable.
|
|
|
|
Source : AIMsight Station Guide, rév. 1.5
|