feat(ocr-pipeline): add Step 5 word recognition (grid from columns × rows)

Backend: build_word_grid() intersects column regions with content rows,
OCRs each cell with language-specific Tesseract, and returns vocabulary
entries with percent-based bounding boxes. New endpoints: POST /words,
GET /image/words-overlay, ground-truth save/retrieve for words.
Frontend: StepWordRecognition with overview + step-through labeling modes,
goToStep callback for row correction feedback loop.
MkDocs: OCR Pipeline documentation added.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

admin-lehrer/app/(admin)/ai/ocr-pipeline/page.tsx

@@ -112,6 +112,16 @@ export default function OcrPipelinePage() {
     }
   }
 
+  const goToStep = (step: number) => {
+    setCurrentStep(step)
+    setSteps((prev) =>
+      prev.map((s, i) => ({
+        ...s,
+        status: i < step ? 'completed' : i === step ? 'active' : 'pending',
+      })),
+    )
+  }
+
   const handleNext = () => {
     if (currentStep < steps.length - 1) {
       setSteps((prev) =>
@@ -161,7 +171,7 @@ export default function OcrPipelinePage() {
       case 3:
         return <StepRowDetection sessionId={sessionId} onNext={handleNext} />
       case 4:
-        return <StepWordRecognition />
+        return <StepWordRecognition sessionId={sessionId} onNext={handleNext} goToStep={goToStep} />
       case 5:
         return <StepCoordinates />
       case 6:

admin-lehrer/app/(admin)/ai/ocr-pipeline/types.ts

@@ -29,6 +29,7 @@ export interface SessionInfo {
   dewarp_result?: DewarpResult
   column_result?: ColumnResult
   row_result?: RowResult
+  word_result?: WordResult
 }
 
 export interface DeskewResult {
@@ -116,6 +117,46 @@ export interface RowGroundTruth {
   notes?: string
 }
 
+export interface WordBbox {
+  x: number
+  y: number
+  w: number
+  h: number
+}
+
+export interface WordEntry {
+  row_index: number
+  english: string
+  german: string
+  example: string
+  confidence: number
+  bbox: WordBbox
+  bbox_en: WordBbox | null
+  bbox_de: WordBbox | null
+  bbox_ex: WordBbox | null
+  status?: 'pending' | 'confirmed' | 'edited' | 'skipped'
+}
+
+export interface WordResult {
+  entries: WordEntry[]
+  entry_count: number
+  image_width: number
+  image_height: number
+  duration_seconds: number
+  summary: {
+    total_entries: number
+    with_english: number
+    with_german: number
+    low_confidence: number
+  }
+}
+
+export interface WordGroundTruth {
+  is_correct: boolean
+  corrected_entries?: WordEntry[]
+  notes?: string
+}
+
 export const PIPELINE_STEPS: PipelineStep[] = [
   { id: 'deskew', name: 'Begradigung', icon: '📐', status: 'pending' },
   { id: 'dewarp', name: 'Entzerrung', icon: '🔧', status: 'pending' },