Add double-page spread detection to frontend pipeline
Some checks failed
CI / go-lint (push) Has been skipped
CI / python-lint (push) Has been skipped
CI / nodejs-lint (push) Has been skipped
CI / test-go-school (push) Successful in 36s
CI / test-go-edu-search (push) Successful in 34s
CI / test-python-klausur (push) Failing after 2m0s
CI / test-python-agent-core (push) Successful in 17s
CI / test-nodejs-website (push) Successful in 18s

After orientation detection, the frontend now automatically calls the
page-split endpoint. When a double-page book spread is detected, two
sub-sessions are created and each goes through the full pipeline
(deskew/dewarp/crop) independently — essential because each page of a
spread tilts differently due to the spine.

Frontend changes:
- StepOrientation: calls POST /page-split after orientation, shows
  split info ("Doppelseite erkannt"), notifies parent of sub-sessions
- page.tsx: distinguishes page-split sub-sessions (current_step < 5)
  from crop-based sub-sessions (current_step >= 5). Page-split subs
  only skip orientation, not deskew/dewarp/crop.
- page.tsx: handleOrientationComplete opens first sub-session when
  page-split was detected

Backend changes (orientation_crop_api.py):
- page-split endpoint falls back to original image when orientation
  rotated a landscape spread to portrait
- start_step parameter: 1 if split from original, 2 if from oriented

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Benjamin Admin
2026-03-24 11:09:44 +01:00
parent 40815dafd1
commit 247b79674d
3 changed files with 115 additions and 19 deletions

View File

@@ -1,19 +1,29 @@
'use client'
import { useCallback, useEffect, useState } from 'react'
import type { OrientationResult, SessionInfo } from '@/app/(admin)/ai/ocr-pipeline/types'
import type { OrientationResult, SessionInfo, SubSession } from '@/app/(admin)/ai/ocr-pipeline/types'
import { ImageCompareView } from './ImageCompareView'
const KLAUSUR_API = '/klausur-api'
interface PageSplitResult {
multi_page: boolean
page_count?: number
sub_sessions?: { id: string; name: string; page_index: number }[]
used_original?: boolean
duration_seconds?: number
}
interface StepOrientationProps {
sessionId?: string | null
onNext: (sessionId: string) => void
onSubSessionsCreated?: (subs: SubSession[]) => void
}
export function StepOrientation({ sessionId: existingSessionId, onNext }: StepOrientationProps) {
export function StepOrientation({ sessionId: existingSessionId, onNext, onSubSessionsCreated }: StepOrientationProps) {
const [session, setSession] = useState<SessionInfo | null>(null)
const [orientationResult, setOrientationResult] = useState<OrientationResult | null>(null)
const [pageSplitResult, setPageSplitResult] = useState<PageSplitResult | null>(null)
const [uploading, setUploading] = useState(false)
const [detecting, setDetecting] = useState(false)
const [error, setError] = useState<string | null>(null)
@@ -92,13 +102,38 @@ export function StepOrientation({ sessionId: existingSessionId, onNext }: StepOr
corrected: orientData.corrected,
duration_seconds: orientData.duration_seconds,
})
// Auto-trigger page-split detection (double-page book spreads)
try {
const splitRes = await fetch(
`${KLAUSUR_API}/api/v1/ocr-pipeline/sessions/${data.session_id}/page-split`,
{ method: 'POST' },
)
if (splitRes.ok) {
const splitData: PageSplitResult = await splitRes.json()
setPageSplitResult(splitData)
if (splitData.multi_page && splitData.sub_sessions && onSubSessionsCreated) {
onSubSessionsCreated(
splitData.sub_sessions.map((s) => ({
id: s.id,
name: s.name,
box_index: s.page_index,
current_step: splitData.used_original ? 1 : 2,
}))
)
}
}
} catch (e) {
console.error('Page-split detection failed:', e)
// Not critical — continue as single page
}
} catch (e) {
setError(e instanceof Error ? e.message : 'Unbekannter Fehler')
} finally {
setUploading(false)
setDetecting(false)
}
}, [sessionName])
}, [sessionName, onSubSessionsCreated])
const handleDrop = useCallback((e: React.DragEvent) => {
e.preventDefault()
@@ -225,6 +260,29 @@ export function StepOrientation({ sessionId: existingSessionId, onNext }: StepOr
</div>
)}
{/* Page-split result */}
{pageSplitResult?.multi_page && (
<div className="bg-blue-50 dark:bg-blue-900/20 rounded-lg border border-blue-200 dark:border-blue-700 p-4">
<div className="text-sm font-medium text-blue-700 dark:text-blue-300">
Doppelseite erkannt {pageSplitResult.page_count} Seiten
</div>
<p className="text-xs text-blue-600 dark:text-blue-400 mt-1">
Jede Seite wird einzeln durch die Pipeline (Begradigung, Entzerrung, Zuschnitt, ...) verarbeitet.
{pageSplitResult.used_original && ' (Seitentrennung auf dem Originalbild, da die Orientierung die Doppelseite gedreht hat.)'}
</p>
<div className="flex gap-2 mt-2">
{pageSplitResult.sub_sessions?.map((s) => (
<span
key={s.id}
className="text-xs px-2 py-1 rounded-md bg-blue-100 dark:bg-blue-800/40 text-blue-700 dark:text-blue-300"
>
{s.name}
</span>
))}
</div>
</div>
)}
{/* Next button */}
{orientationResult && (
<div className="flex justify-end">
@@ -232,7 +290,7 @@ export function StepOrientation({ sessionId: existingSessionId, onNext }: StepOr
onClick={() => onNext(session.session_id)}
className="px-6 py-2 bg-teal-600 text-white rounded-lg hover:bg-teal-700 font-medium transition-colors"
>
Weiter &rarr;
{pageSplitResult?.multi_page ? 'Seiten verarbeiten' : 'Weiter'} &rarr;
</button>
</div>
)}