Help
Glossary
Plain-language definitions of the terms you’ll see across Scanix.
A quick reference for the words that show up across the app and these docs. Each entry is one or two sentences with a link to the page that covers it in full. Definitions match the exact labels Scanix uses on screen.
A–C
| Term | What it means |
|---|---|
| AI capture | Reading a document by handing the whole page to a model that classifies it and returns structured fields, instead of drawing fixed zones. The flexible counterpart to a template. See Templates vs AI capture. |
| AI module | A pre-built extraction schema for one document class (for example Invoices AI or Identity Documents AI) that tells the AI which fields to pull and normalises the result. Set one on a Job Group or template. See How Scanix works. |
| AI Service | A document-type capture module you switch on under Settings → AI Services. Active services appear in the Job Group and Template designers and drive AI document splitting and extraction; a deactivated service is skipped at extraction time (logged as a per-document skip rather than failing the batch). |
| Analyse with AI | The primary Viewer button that appears when an AI service is selected. One click runs OCR and the AI module together and fills your mapped fields. See Templates vs AI capture. |
| Anchor / Anchoring | Pins a zone to a nearby landmark — a Text label, Barcode, QR code, or Page edge — so Scanix re-positions the zone when a scan drifts, shifts, skews, or rotates. Set per zone in the Designer. |
| Auto-accept | The confidence bar (Auto-accept above N% confidence, default 97%) above which a document's fields export automatically; anything below is held in the Verification Queue. Set under Settings → AI Services. |
| Barcode / QR Code | A 1D or 2D code Scanix can read as a captured field value, or use as a document separator. Supported symbologies include QR Code, Code 128, Code 39, EAN-13, Data Matrix, PDF417, and Aztec. See Barcodes & QR codes. |
| Batch | A set of documents captured together in one pass — a scan, an import, or a folder drop — that move through the pipeline as a group. |
| Capture | The first pipeline stage: bringing documents in by scanning, importing files, drag-and-drop, importing a folder, or a watched Hot Folder. See How Scanix works. |
| Catch-all | A member template whose matcher claims every document it's offered. It belongs at the bottom of a Job Group's priority list as a fallback — placed higher, it shadows everything below it. |
| Check digit | A digit built into an MRZ (per the ICAO 9303 standard) that Scanix recomputes from the characters it read to catch a misread. A mismatch sends the field to review rather than discarding it. |
| Classification | Deciding which type a document is — invoice, receipt, passport — so it can be routed to the right template or AI module. Done by member matchers or by AI. |
| Compression (MRC) | A per-output level on a Searchable PDF that shrinks the file using mixed-raster compression. Levels run Off — no compression, Low, Medium, High (recommended), Extra High, and Archival — lossless text. Exposed only after you tick Enable high-compression document output (MRC) in Settings → General. |
D–I
| Term | What it means |
|---|---|
| DPI | Dots per inch — the scan resolution. The Resolution dropdown offers 100 / 150 / 200 / 300 / 400 / 600 DPI (trimmed to what your scanner advertises); the free-form Scan options dialog defaults to 200 DPI, and a template defaults to 300 (recommended). Higher resolution costs speed and disk on every page. |
| Document splitting | Breaking a stack into separate documents using a rule — a Barcode separator, a patch-code sheet, or blank pages — so the batch files as individual documents. See How Scanix works. |
| Duplex | Capturing the front and back of each sheet in one pass. Needs the ADF / Feeder paper source — the Duplex (scan both sides — feeder only) checkbox is greyed out on flatbed and simplex-only scanners. |
| Enhance | The pipeline stage that cleans up the page image — straightening, cropping, and colour handling — configured per template before OCR runs. See How Scanix works. |
| Export | The final pipeline stage: writing results to the formats and destinations you pick, such as JSON, CSV, or a Searchable PDF. |
| Hot Folder | A watched folder on disk that automatically imports, reads, and (in Auto-process mode) files any document dropped into it — a lights-out pipeline. A higher-tier feature. |
| Index field | A named column the captured values land in when you export — the metadata for a document. Created from zones (Import from Zones) or by hand, and mapped from OCR zones, MRZ, barcodes, or AI. See Index fields. |
J–O
| Term | What it means |
|---|---|
| Job Group | A container that ties several member templates together and routes each document in a mixed batch to the right one in a single pass — split, route, OCR, extract, export. See Job Groups. |
| Member (template) | A template added to a Job Group. Members form a priority list; the first whose splitting rule recognises a document claims it (first match wins). |
| MRZ | The machine-readable zone — the band of <-padded uppercase text at the bottom of a passport or ID card. Set a zone's Type to MRZ and Scanix decodes it into named, check-digit-verified fields (Surname, Date of Birth, Nationality, and more). See Passport & ID (MRZ). |
| OCR | Optical character recognition — reading the printed text on a page. Runs on your device, offline once language packs have downloaded. The engine is chosen under Settings → OCR Engine. |
| Output | A template's export configuration — a Document Format for the pages, a Metadata Format for the index data, and a destination. Either format can be set to None. |
P–R
| Term | What it means |
|---|---|
| Per-page capture | Extraction that records one value per page for each field, tagged with the page it came from, and carries that page number through the on-screen field list and into every export. See Per-page fields. |
| Pipeline | The five stages every document moves through, in order: Capture → Enhance → OCR / AI → Verify → Export. See How Scanix works. |
| Privacy mode (Local-only) | The Local-only mode (no cloud egress) control under Settings → AI Services — a global hard stop that blocks every AI service from sending document content to a cloud provider, regardless of any per-group or per-template routing. A per-template or per-group equivalent (Privacy mode: local only) lives on the designer's AI Services tab. |
| Required field | An index field a document shouldn't pass without. Mark it Required (or the stricter Mandatory, which can't be overridden) in the Designer; export is held back while the value is missing, and the run tells you which pages to fix. See Index fields. |
| Routing | Deciding which member template owns each document in a Job Group, by first match wins down the priority order. Unmatched documents go to an Unrecognized (or Unclassified) review pile. See How routing works. |
| Routing policy | How an AI job picks where to run: Local first, cloud fallback, Local only, Cloud only, or Pinned to one provider. Set the default under Settings → AI Services; a Job Group or template can pin its own. |
S–Z
| Term | What it means |
|---|---|
| Searchable PDF | A PDF with an invisible OCR text layer, so the document is full-text searchable. The default Document Format, and the only one that supports multi-document filing and high-compression. |
| Separator | A page that marks where one document ends and the next begins — most often a barcode or patch-code sheet. Optionally removed from the finished output. |
| Smart Template | A capture template that AI builds from a single sample page — it proposes the document type, zones, index fields, splitting strategy, outputs, and (when confident) an anchor. See Templates vs AI capture. |
| Symbology | The type of barcode — QR Code, Code 128, Code 39, EAN-13, Data Matrix, PDF417, Aztec. Picked per zone or per splitting rule. |
| Template | A saved recipe for one document type that stores the enhancement, OCR, zone, and export settings, so the same kind of document is always processed the same way. Built in the Custom Builder (the Designer). See Templates vs AI capture. |
| Type (zone) | The setting on a template zone that decides how its value is read and cleaned — Text, Number, Currency, Date, Email, Phone, MRZ, or a code type such as Barcode or QR Code. |
| Verify / Verification Queue | The pipeline stage where a person reviews extracted fields before export. Scanix marks each field valid, needs-review, or invalid; documents below the Auto-accept bar are held here. See How Scanix works. |
| Zone | A rectangle you draw on a template's sample page. On every incoming page Scanix finds that same region and reads only what's inside it; the zone's Type decides how the value is read. See How Scanix works. |
Required vs Mandatory
A field can be marked Required or Mandatory in the Designer (shown as Req. in the field table). Both block export while the value is empty — but a Required field lets an operator confirm and export anyway, while a Mandatory field can't be overridden and must be filled first. An optional field that read nothing simply drops out of that page's field list. See Index fields.
Capture types at a glance
The fastest way to think about reading structured data is by a zone's Type.
| Type | What it captures |
|---|---|
| Text, Number, Currency, Date, Email, Phone | A plain region — Scanix reads the value and tidies it, stripping a printed label if the box caught one. |
| MRZ | A passport or ID band, decoded into separate check-digit-verified fields. |
| Barcode, QR Code, PDF417, Aztec, Data Matrix | A 1D or 2D code, decoded into the field's value. |
Image-enhancement terms vary by build
Names for individual image-cleanup steps (deskew, binarise, colour dropout, colour mode) differ between builds and aren't fixed labels in these docs. Configure them per template in the Designer's enhancement settings and preview the result on your sample page rather than relying on a specific label here.
Screenshot
A two-column reference layout of the glossary — alphabetised term on the left, one-line definition on the right — with the “Capture types at a glance” table below it. — shot glossary-01
Next steps
How Scanix works
The Capture → Enhance → OCR / AI → Verify → Export pipeline, and a tour of the app.
FAQ
Short answers to the questions operators ask most.