DOT Web Document Auto Capture



DOT Web Document Auto Capture is set of Non-ui component and Ui component web components to capture image of an ID document.

Non-ui component is a web component that renders the video stream from an available phone or web camera to automatically capture an image of an ID document with the required quality.

Ui component is a web component that renders an overlay over video stream. Overlay includes a placeholder, camera control buttons and instructions to guide the user to position the document correctly.

Supported Browsers

DOT Web Document Auto Capture was tested with:

  • Chrome on desktop (Windows, Mac and Linux) and mobile (Android, iPhone)

  • Firefox on desktop (Windows, Mac and Linux) and mobile (Android)

  • Edge on Windows

  • Safari on Mac, iPad and iPhone

  • WebView on Android*

  • SafariVC on iPad and iPhone**

Known issues:

  • Components don’t work with WKWebView on iPhone

* Some older Android phones can experience issues with camera stream initialization using WebRTC. Our web components will work on most of them, but camera stream could be slow and laggy. This issue is caused by device and can be experienced in any website using WebRTC in Android WebView.

** Components are tested with SafariVC on devices iPhone 7 and newer with iOS 15 or iPadOS 15 or newer. Older devices or iOS versions are not officially supported and might experience issues.

Privacy and security

This component can only be used in secure contexts due to MediaDevices API used for handling camera access. A secure context is, in short, a page loaded using HTTPS or the file:/// URL scheme, or a page loaded from localhost. Before accessing any camera, the component must always get the user’s permission. Browsers may offer a once-per-domain permission feature, but they must ask at least the first time, and the user has to specifically grant ongoing permission if they choose to do so. Browsers are required to display an indicator that shows that a camera or microphone is in use. More details can be found on MDN docs.

Non-ui component


Minimum required camera resolution for appropriate results is 720p. Anything less than 720p is insufficient.


DOT Web Document Auto Capture can be installed via NPM, yarn or pnpm

npm install @innovatrics/dot-document-auto-capture

To manually integrate the DOT Web Document Auto Capture, download latest version from the Github repository. Add following line to dependencies in your package.json:

"dependencies": {
    "@innovatrics/dot-document-auto-capture": "file:dot-document-auto-capture-[VERSION].tgz",

where [VERSION] is the DOT Web Document Auto Capture version integrated. This installs dot-document-auto-capture as an external module that can be use then (just like any other module in the code) For example, one could do import '@innovatrics/dot-document-auto-capture'; in the app.


Document auto capture component is a web component which uses custom HTML <x-dot-document-auto-capture/> tag. Properties cameraOptions needs to be passed into component after <x-dot-document-auto-capture/> tag was rendered.

import type {
} from '@innovatrics/dot-document-auto-capture';
import { useCallback, useEffect } from 'react';
import '@innovatrics/dot-document-auto-capture';

const DocumentCamera = (props: DocumentCameraProps) => {
  useEffect(() => {
   const documentAutoCaptureHTMLElement = document.getElementById(
    ) as HTMLDocumentCaptureElement | null;

    if (documentAutoCaptureHTMLElement) {
      documentAutoCaptureHTMLElement.cameraOptions = props;

 return <x-dot-document-auto-capture id="x-dot-document-auto-capture" />;

const Page = () => {

  const handleDocumentPhotoTaken = (image: Blob, data: DocumentComponentData) => {
    // ...

  // Save function reference to prevent unnecessary reload of component
  const handleError = useCallback(
    (error: Error) => {

  return (


Declaration file is bundled inside package. To use with TypeScript, import types from @innovatrics/dot-document-auto-capture.

import type { DocumentCallback, DocumentCameraProps } from '@innovatrics/dot-document-auto-capture';

Hosting SAM wasm

The component needs to have access to the WebAssembly wasm binary file. It’s distributed in the package and needs to be hosted by the website provider. By default, the component will try to fetch the wasm file from <PROJECT_ORIGIN>/sam.wasm. This can by changed using samWasmUrl property. If using Create React App, copy sam.wasm file to public folder. In our example, the final path is public/sam.wasm.


  • (Optional) ['png'] string imageType – Format of the image returned after successful capture

    • 'jpeg'

    • 'png'

  • (Optional) string cameraFacing – Defines which camera to acquire from browser’s getUserMedia API. Default camera facing for mobile phones is set to environment and for others platforms is set to user

    • 'user' – The video source is facing toward the user; this is the selfie or front-facing camera on a smartphone

    • 'environment' – The video source is facing away from the user, thereby viewing their environment; this is the back camera on a smartphone

  • function onPhotoTaken – Callback on successful image capture

  • function onError – Callback for the case that an error occurred (see Handling errors)

    • (e: Error) ⇒ void

  • (Optional) string samWasmUrl - URL link to the location where the wasm binary file is hosted

  • (Optional) ['standard'] string validationMode – Sets the validation mode for document placement

    • 'standard' – Document can be placed anywhere, but cannot leak out of photo

    • 'strict' – Document must be placed inside placeholder

  • (Optional) object thresholds - Detection configuration

    • (Optional) [0.8] number confidenceThreshold - Detection confidence threshold

    • (Optional) [250] number brightnessLowThreshold - Low brightness threshold

    • (Optional) [900] number brightnessHighThreshold - High brightness threshold

    • (optional) [0.03] number outOfBoundsThreshold - Document out of bounds threshold

    • (Optional) [500] number sharpnessThreshold - Low sharpness threshold

    • (Optional) [100] number hotspotsScoreThreshold - Hotspots score threshold

    • (Optional) [0.43] number sizeSmallThreshold - Small size threshold

Callback parameters

  • Blob image – Returned image on successful capture

  • object data

    • object cameraSettings - MediaTrackSettings object containing used webcam settings. The object in addition contains device name.

    • (Optional) object detection - Object contains all detection parameters and its values. Present if image was taken using auto capture (not manual capture).

    • object imageResolution - Width and height of the captured image.

Multi capture

Document auto capture component allows you to capture an unlimited number of documents without the need to reinitialize the webcam and detector. This allows you to capture two sides of a document or multiple documents. Component calls onPhotoTaken callback on every captured document photo. When onPhotoTaken is called, detection is paused. Camera stream and document detector stay initialized. Component is in waiting state. You should implement a custom UI for waiting state (e.g. overlay over our component). We provide a default UI, but it is not customizable.

To continue detection, dispatch continue-detection event.

Implement dispatch 'continue-detection' event

import { dispatchControlEvent, DocumentCustomEvent, ControlEventInstruction } from '@innovatrics/dot-document-auto-capture/events';

const continueDetection = () => {
  dispatchControlEvent(DocumentCustomEvent.CONTROL, ControlEventInstruction.CONTINUE_DETECTION)

Without importing @innovatrics/dot-document-auto-capture/events.

function continueDetection() {
    new CustomEvent('document-auto-capture:control', {
      detail: { instruction: 'continue-detection' },

Handling errors

When an error occurs we call onError callback with one parameter of type Error. We set name property to AutoCaptureError and also message with more details. Component renders default UI for error state but is not customizable, and integrator should implement own error handling. Component uses the MediaDevices API that provides access to connected media input devices like cameras and microphones. If the user denies permission to access or the webcam is not present, an error is thrown. We provide original error thrown by browser inside cause property of the error object. List of possible errors can be found on MDN docs.

Error example:

  name: "AutoCaptureError",
  message: "The webcam is already in use by another application",
  cause: DOMException: Could not start video source // Original error thrown by browser MediaDevices API

Ui component


Both components must be wrapped in parent div with position: relative.

<div style={{position: 'relative'}}>
  <DocumentUi />


UI component can be installed via NPM, yarn or pnpm

npm install @innovatrics/dot-auto-capture-ui

To manually integrate UI component, download latest version from the Github repository. Add following line to dependencies in your package.json:

"dependencies": {
    "@innovatrics/dot-auto-capture-ui": "file:dot-auto-capture-ui-[VERSION].tgz",

where [VERSION] is the DOT Web Document Auto Capture version integrated. This installs dot-auto-capture-ui as an external module that can be used (just like any other module in the code). For example, one could do import '@innovatrics/dot-auto-capture-ui'; in the app.


Document auto capture UI component is an web component which uses custom HTML <x-dot-document-auto-capture-ui /> tag. Properties props needs to be passed into component after <x-dot-document-auto-capture-ui /> tag was rendered.

import type { DocumentUiProps, HTMLDocumentUiElement } from '@innovatrics/dot-auto-capture-ui/document'
import '@innovatrics/dot-auto-capture-ui/document';
import { useEffect } from 'react';

const DocumentUi = (props: DocumentUiProps) => {
  useEffect(() => {
    const uiElement = document.getElementById('x-dot-document-auto-capture-ui') as HTMLDocumentUiElement | null;

    if (uiElement) {
      uiElement.props = props;

  return <x-dot-document-auto-capture-ui id="x-dot-document-auto-capture-ui" />;


Declaration files are bundled inside package. Just import types from @innovatrics/dot-auto-capture-ui/document.


  • (Optional) enum placeholder - One of the predefined placeholders in component that can be selected

    • ID_CORNERS = 'id-corners'

    • ID_DASH = 'id-dash'

    • ID_DOT = 'id-dot'

    • ID_SOLID = 'id-solid'

    • ID_PHOTO_ROUNDED = 'id-photo-rounded'

    • ID_CORNERS_ROUNDED = 'id-corners-rounded'

    • ID_DASH_ROUNDED = 'id-dash-rounded'

    • ID_DOT_ROUNDED = 'id-dot-rounded'

    • ID_SOLID_ROUNDED_BACK = 'id-solid-rounded-back'

    • ID_SOLID_ROUNDED = 'id-solid-rounded'

    • PASSPORT_SOLID_BACK = 'passport-solid-back'

    • PASSPORT_SOLID_BACK_BLANK = 'passport-solid-back-blank'

  • (Optional) object instructions - Modification of default messages for localization or customization

    • (Optional) ['Hold still…'] string candidate_selection - Shown when all validations are passed, i.e. image is suitable for capture

    • (Optional) ['Place document in rectangle'] string document_centering - Shown when the document is not centered inside the placeholder

    • (Optional) ['Move back'] string document_too_close - Shown when the document is too close to the camera

    • (Optional) ['Place document in rectangle'] string document_not_present - Shown when no document is detected

    • (Optional) ['Move closer'] string document_too_far - Shown when the document is too far from the camera.

    • (Optional) ['More light needed'] string sharpness_too_low - Shown when the document found in the image is not sharp enough

    • (Optional) ['More light needed'] string brightness_too_low - Shown when the image is too dark

    • (Optional) ['Less light needed'] string brightness_too_high - Shown when the image is too bright.

    • (Optional) ['Avoid reflections'] string hotspots_present - Shown when the document found in the image has reflections

  • (Optional) theme - Modification of theme properties.

    • (Optional) colors - To customize colors of following properties

      • (Optional) ['white'] color placeholderColor - Color of the placeholder lines

      • (Optional) ['#00BFB2'] color placeholderColorSuccess - Color of the placeholder lines when all validations are passed

      • (Optional) ['white'] color instructionColor - Instruction background color

      • (Optional) ['#00BFB2'] color instructionColorSuccess - Instruction background color when all validations are passed

      • (Optional) ['black'] color instructionTextColor - Instruction text color

  • (Optional) appStateInstructions - Modification of default messages for component state

    • (Optional) loading - Component loading state

      • (Optional) ['Loading. Please wait.'] string text - Text shown while component is loading

      • (Optional) [true] boolean visible - Show/hide loading instruction while component is loading

    • (Optional) waiting - Component waiting state

      • (Optional) ['Waiting for input'] string text - Text shown while component is waiting

      • (Optional) [true] boolean visible - Show/hide waiting instruction while component is waiting

  • (Optional) [false] boolean showDetectionLayer - Show/hide detection layer on top of component

  • (Optional) [false] boolean showCameraButtons - Show/hide camera buttons (switch camera and toggle mirror) on top of component

Example how to use UI properties:

import { DocumentPlaceholderIcon } from '@innovatrics/dot-auto-capture-ui/document';

  theme={{ colors: { placeholderColor: 'green' } }}
  instructions={{ candidate_selection: 'Candidate selection' }}
  appStateInstructions={{ loading: { text: 'Component is loading', visible: true } }}

In javascript you can use values of DocumentPlaceholderIcon directly.

<DocumentUi placeholder="id-dash" />

Custom Events - Optional

Communication between components is based on custom events. Ui component is mostly listening to events dispatched by Non-ui component. In case of CONTROL event, Ui component dispatches control events to control Non-ui component.

When using default Ui component, only dispatch continue-detection event has to be implemented to use Multi capture.

Type of Events

Currently all events being used are described in this section. These are sufficient to build custom UI layer.

  • enum DocumentCustomEvent - Event names dispatched by components

    • CONTROL = 'document-auto-capture:control' - Events dispatched from Ui component to control Non-ui component. Described in Control Events section

    • CAMERA_PROPS_CHANGED = 'document-auto-capture:camera-props-changed' - Notifies UI when camera properties has changed

    • INSTRUCTION_CHANGED = 'document-auto-capture:instruction-changed' - Notifies UI when instruction has changed

    • STATE_CHANGED = 'document-auto-capture:state-changed' - Notifies UI when state of Non-ui component has changed

    • DETECTED_DOCUMENT_CHANGED = 'document-auto-capture:detected-document-changed' - Notifies UI when document corners are detected. Used in Detection Layer, has no effect if showDetectionLayer = false

    • VIDEO_ELEMENT_SIZE = 'document-auto-capture:video-element-size' - Notifies UI when HTML video element size has changed


Import @innovatrics/dot-document-auto-capture/events to use event types and and dispatch control events.

Listening to the Events

All event listeners are already implemented in default Ui component. Skip this section if you are using default Ui component.

All DocumentCustomEvent events, except CONTROL, are dispatched by Document Auto Capture Non-ui component. Ui component listens to these events to make appropriate changes. See the example below how to register event listeners.

import type {
} from '@innovatrics/dot-document-auto-capture/events'
import { DocumentCustomEvent } from '@innovatrics/dot-document-auto-capture/events';
import type {
} from '@innovatrics/dot-document-auto-capture';
import { useEffect, useState } from 'react'

const Events = () => {
  const [instructionCode, setInstructionCode] = useState<DocumentInstructionCode | undefined>();
  const [cameraResolution, setCameraResolution] = useState<Resolution | undefined>();
  const [isMirroring, setIsMirroring] = useState<boolean | undefined>();
  const [appState, setAppState] = useState<AppState | undefined>();
  const [error, setError] = useState<AutoCaptureError | undefined>();
  const [detectedDocumentCorners, setDetectedDocumentCorners] = useState<DetectedDocumentCorners | undefined>();
  const [videoElementSize, setVideoElementSize] = useState<DOMRect | undefined>();

  useEffect(() => {
    const handleInstruction = (event: DocumentInstructionChangeEvent) => {

    const handleCameraProps = (event: CameraPropsChangeEvent) => {

    const handleAppState = (event: CameraStateChangeEvent) => {

      const error = event?.detail?.error;

      if (error) {

    const handleDetectedDocument = (event: DetectedDocumentChangeEvent) => {

    const handleVideoElementSize = (event: VideoElementSizeChangeEvent) => {

    document.addEventListener(DocumentCustomEvent.INSTRUCTION_CHANGED, handleInstruction);
    document.addEventListener(DocumentCustomEvent.CAMERA_PROPS_CHANGED, handleCameraProps);
    document.addEventListener(DocumentCustomEvent.STATE_CHANGED, handleAppState);
    document.addEventListener(DocumentCustomEvent.DETECTED_DOCUMENT_CHANGED, handleDetectedDocument);
    document.addEventListener(DocumentCustomEvent.VIDEO_ELEMENT_SIZE, handleVideoElementSize)

    return () => {
      document.removeEventListener(DocumentCustomEvent.INSTRUCTION_CHANGED, handleInstruction);
      document.removeEventListener(DocumentCustomEvent.CAMERA_PROPS_CHANGED, handleCameraProps);
      document.removeEventListener(DocumentCustomEvent.STATE_CHANGED, handleAppState);
      document.removeEventListener(DocumentCustomEvent.DETECTED_DOCUMENT_CHANGED, handleDetectedDocument);
      document.removeEventListener(DocumentCustomEvent.VIDEO_ELEMENT_SIZE, handleVideoElementSize);
  }, [])

Without importing @innovatrics/dot-document-auto-capture/events you can use values of DocumentCustomEvent directly.

const instructionChangeEvent = "document-auto-capture:instruction-changed";

function handleInstructionChange(event) {

document.addEventListener(instructionChangeEvent, handleInstructionChange);

 * remove event listener when you're done
document.removeEventListener(instructionChangeEvent, handleInstructionChange);

Control Events

Control events are dispatched from Ui component to control Non-ui component.

  • enum ControlEventInstruction

    • CONTINUE_DETECTION = 'continue-detection' - Controls Multi capture

    • SWITCH_CAMERA = 'switch-camera' - Notifies Document Auto Capture to use different camera

    • TOGGLE_MIRROR = 'toggle-mirror' - Notifies Document Auto Capture to mirror video stream

import { dispatchControlEvent, DocumentCustomEvent, ControlEventInstruction } from '@innovatrics/dot-document-auto-capture/events';

export const Dispatch = () => {

  const continueDetection = () => {
    dispatchControlEvent(DocumentCustomEvent.CONTROL, ControlEventInstruction.CONTINUE_DETECTION)

  const switchCamera = () => {
    dispatchControlEvent(DocumentCustomEvent.CONTROL, ControlEventInstruction.SWITCH_CAMERA)

  const toggleMirror = () => {
    dispatchControlEvent(DocumentCustomEvent.CONTROL, ControlEventInstruction.TOGGLE_MIRROR)

  return (
      <button onClick={continueDetection}>Continue detection</button>
      <button onClick={switchCamera}>Switch camera</button>
      <button onClick={toggleMirror}>Mirror camera</button>

Without importing @innovatrics/dot-document-auto-capture/events you can use values of DocumentCustomEvent and ControlEventInstruction directly.

const continueDetection = () => {
    new CustomEvent('document-auto-capture:control', {
      detail: { instruction: 'continue-detection' },

Example of using components together

import type { DocumentComponentData } from '@innovatrics/dot-document-auto-capture';
import {
} from '@innovatrics/dot-document-auto-capture/events'
import { useCallback } from 'react';

import DocumentCamera from './DocumentCamera'
import DocumentUi from './DocumentUi'

const DocumentAutoCapture = () => {
  const handlePhotoTaken = (image: Blob, data: DocumentComponentData) => {
    console.log(image, data);

  const handleError = useCallback(
    (error: Error) => {

  const handleContinueDetection = () => {
    dispatchControlEvent(DocumentCustomEvent.CONTROL, ControlEventInstruction.CONTINUE_DETECTION)

  return (
      <button onClick={handleContinueDetection}>Continue detection</button>
      <div style={{ position: 'relative' }}>
        <DocumentUi />

Code Samples

See also DOT Web Samples showing the usage of DOT Web Auto Capture components in different front-end technologies like React, Angular…​



4.0.1 - 2023-01-19

  • export events from @innovatrics/dot-document-auto-capture/events

4.0.0 - 2023-01-19

  • UI layer from camera component (buttons, placeholder, instruction and app state overlay)

  • strict mode and placeholder fit validator

  • UI package with placeholder, instructions, buttons and detection layer

  • Custom events and events listeners for communication from app to component (switch-camera, toggle-mirror)

  • Custom events and events listeners for communication from components to app (camera-resolution-changed, instruction-changed, detected-document-changed, state-changed)

  • showCameraButtons switch for show/hide camera buttons (switch camera, toggle mirror)

  • SAM version into debug layer

  • Rename photoTakenCb to onPhotoTaken

  • onPhotoTaken returns full frame image, instead of cropped

  • Order of instruction processing. Before validating the centering of the detected document, the size of the detected document is validated.

  • Component is now build as umd and es module

  • Camera buttons (switch camera, toggle mirror) are now hidden by default

3.5.2 - 2022-12-16

  • Camera switch on Android devices with Firefox

3.5.1 - 2022-10-13


Memorized props from first instance which holds reference to callback function from initialization on every capture

3.5.0 - 2022-10-10

  • Dynamic candidate selection phase duration

  • Minimal camera resolution check

  • Selection of best image in candidate selection phase entering

3.4.3 - 2022-09-28

  • Possible false candidate selection on first frame after entering candidate selection

3.4.2 - 2022-08-10

  • Typescript declaration files export

3.4.1 - 2022-08-09

  • Typescript declaration files export

  • Dependencies update and upgrade

3.3.1 - 2022-07-04

  • Zero height of component when error occurs

  • add appStateInstructions property to uiCustomisation

  • add option to change appState instructions text and visibility

  • interface for changing Loading appState instruction text

3.3.0 - 2022-06-03

  • Parent of <x-dot-document-auto-capture /> does not have to have a defined height

3.2.0 - 2022-05-13

  • Add support for capturing multiple document photos

  • Add custom event document-auto-capture with continue-detection instruction to continue detection

  • Add Countly analytics tracking

  • Session recording for document detection

  • onError callback returns AutoCaptureError

  • onError callback is required

  • Show error screen when error occurs

  • When photoTakenCb is called, component switches into waiting state

  • Stucked component when error occurs

3.1.2 - 2022-04-04

  • Remove force manual capture on instruction click

3.1.0 - 2022-02-14

  • Add detectionLayerVisible with default value false

3.0.3 - 2022-02-03

  • Unify loading and instruction design

3.0.2 - 2022-01-28

  • Add loading screen

  • Add loading.text property in uiCustomisation for configuring text on loading screen

  • Show new instruction only if 600ms elapsed since last instruction change

  • Changed default value for document_centering instruction

3.0.1 - 2022-01-25

  • Introduced possibility to choose between normal and strict validation mode

  • Option to select between normal or strict validation mode.

  • OutOfBoundsValidator and OutOfBoundsThreshold

  • SizeSmallValidator and SizeSmallThreshold

  • safari version 15 on mac

  • Document detection accuracy improved.

  • Detection area size

  • New `sam.wasm`file

  • blaze_models package

  • modelUrls option from cameraOptions

2.3.1 - 2022-01-10

  • Fix bug causing integration error

2.3.0 - 2022-01-05

  • add cameraSettings into photoTakenCb callBack function

  • photoTakenCb callBack function structure

  • Redesign of the UI of Document Auto Capture component.

2.2.2 - 2021-12-17

  • samWasmUrl property for configure path to SAM wasm binary file

2.2.1 - 2021-12-09

  • New SVG placeholder pass-rounded-rectangle-solid-back-blank

  • Hide switch camera button if there is only 1 webcam available

  • Default camera facing for mobile phones is set to environment and for others platforms is set to user

  • Computation of placeholderErrorScore for document placement inside placeholder

  • Default value for placeholderErrorScoreThreshold from 0.03 to 0.035

  • On android phones choose correct default lens if more than one available

2.2.0 - 2021-11-30

  • dynamic import of TensorFlow.js libraries

2.1.0 - 2021-11-18

  • First released version

description: "Face detection and face verification web package. UI and non-UI face components for digital onboarding. Learn more in product documentation."