marcel/familienarchiv
1
0
Fork 0
Code Issues 103 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

refactor(ocr): extract _collapse_adjacent_markers helper and add CORRECTION_MARKER

Browse Source 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Marcel
2026-04-17 16:40:39 +02:00
parent 9a64c0698f
commit 77747aa556
1 changed files with 19 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
ocr-service/confidence.py
View File

@@ -6,6 +6,21 @@ THRESHOLD_DEFAULT = float(os.environ.get("OCR_CONFIDENCE_THRESHOLD", "0.3"))
THRESHOLD_KURRENT = float(os.environ.get("OCR_CONFIDENCE_THRESHOLD_KURRENT", "0.5")) THRESHOLD_KURRENT = float(os.environ.get("OCR_CONFIDENCE_THRESHOLD_KURRENT", "0.5"))
ILLEGIBLE_MARKER = "[unleserlich]" ILLEGIBLE_MARKER = "[unleserlich]"
CORRECTION_MARKER = "[?]"
def _collapse_adjacent_markers(tokens: list[str]) -> list[str]:
collapsed: list[str] = []
prev_was_marker = False
for token in tokens:
if token == ILLEGIBLE_MARKER:
if not prev_was_marker:
collapsed.append(token)
prev_was_marker = True
else:
collapsed.append(token)
prev_was_marker = False
return collapsed
def get_threshold(script_type: str) -> float: def get_threshold(script_type: str) -> float:
@@ -30,19 +45,14 @@ def apply_confidence_markers(words: list[dict], threshold: float | None = None)
if threshold is None: if threshold is None:
threshold = THRESHOLD_DEFAULT threshold = THRESHOLD_DEFAULT
result: list[str] = [] tokens: list[str] = []
prev_was_marker = False
for word in words: for word in words:
if word["confidence"] < threshold: if word["confidence"] < threshold:
if not prev_was_marker: tokens.append(ILLEGIBLE_MARKER)
result.append(ILLEGIBLE_MARKER)
prev_was_marker = True
else: else:
result.append(word["text"]) tokens.append(word["text"])
prev_was_marker = False
return " ".join(result) return " ".join(_collapse_adjacent_markers(tokens))
def words_from_characters(prediction: str, confidences: list[float]) -> list[dict]: def words_from_characters(prediction: str, confidences: list[float]) -> list[dict]: