DOT Android Document library

v3.1.0

Introduction

DOT Android Document provides components for document capture and related functionalities which are easy to integrate into an Android application.

Requirements

DOT Android Document has the following requirements:

Android API level 21

Distribution

Maven Repository

DOT Android Document is distributed as an Android library (.aar package) stored in the Innovatrics maven repository.

In order to integrate DOT Android Document into your project, the first step is to include the Innovatrics maven repository and Google repository to your top level build.gradle file.

build.gradle

allprojects {
    repositories {
        jcenter()
        google()
        maven {
            url 'https://maven.innovatrics.com/releases'
        }
    }
}

Then, specify the dependency on DOT Android Document library in the module’s build.gradle file. Dependencies of this library will be downloaded alongside the library.

build.gradle

dependencies {
    //…
    implementation "com.innovatrics.dot:dot-document:$dotDocumentVersion"
    //…
}

Supported Architectures

DOT Android Document provides binaries for these architectures:

armeabi-v7a
arm64-v8a
x86
x86_64

If your target application format is APK and not Android App Bundle, and the APK splits are not specified, the generated APK file will contain binaries for all available architectures. Therefore we recommend to use APK splits. For example, to generate arm64-v8a APK, add the following section into your module build.gradle:

build.gradle

splits {
    abi {
        enable true
        reset()
        include 'arm64-v8a'
        universalApk false
    }
}

If you do not specify this section, the resulting application can become too large in size.

Permissions

DOT Android Document declares the following permission in AndroidManifest.xml:

AndroidManifest.xml

<uses-permission android:name="android.permission.CAMERA" />

Basic Setup

Logging

Although logging is disabled by default, it can be enabled explicitly by using the following method from the com.innovatrics.android.commons.Logger class.

Logger.setLoggingEnabled(true);

The appropriate place for this call is within the onCreate() method of your subclass of android.app.Application. Each TAG of a log message starts with the prefix dot-document:.

This setting enables logging for all DOT Android libraries.

Keep in mind that logging should be used just for debug purposes as it might produce a lot of log messages.

Components

Overview

DOT Android Document provides both non-UI and UI components. Non-UI components are aimed to be used by developers who want to build their own UI using the DOT Android Document functionality. UI components are build on top of non-UI components. These are available as abstract fragments and can be extended and then embedded into the application’s existing activity providing more control.

List of Non-UI Components

DOCUMENT DETECTOR: A component for performing document detection on an image.
IMAGE PARAMETERS ANALYZER: A component for computing image parameters such as sharpness, brightness or hotspots score.
IMAGE PERSPECTIVE WARPER: A component for perspective warping an image according detected document card corners.
DOCUMENT AUTO CAPTURE CONTROLLER: A component for capturing good quality photos suitable for optical character recognition.
MACHINE READABLE ZONE READER: A component for reading Machine Readable Zone (MRZ).

List of UI Components

DOCUMENT AUTO CAPTURE: A visual component for capturing good quality photos suitable for optical character recognition.

Non-UI Components

Document Detector

The DocumentDetector interface provides a document detection functionality.

Create a DocumentDetector:

DocumentDetector documentDetector = DocumentDetectorFactory.create();

To perform detection, call the following method on the background thread:

DocumentDetector.Result result = documentDetector.detect(bgraRawImage);

Image Parameters Analyzer

The ImageParametersAnalyzer interface provides an image parameters analysis functionality.

Create ImageParametersAnalyzer:

ImageParametersAnalyzer imageParametersAnalyzer = ImageParametersAnalyzerFactory.create();

To perform analysis, call the following method on the background thread:

ImageParameters imageParameters = imageParametersAnalyzer.analyze(bgraRawImage);

Image Perspective Warper

The ImagePerspectiveWarper interface provides perspective warping functionality.

Create ImagePerspectiveWarper:

ImagePerspectiveWarper imagePerspectiveWarper = ImagePerspectiveWarperFactory.create();

To perform perspective warping, call the following method on the background thread:

BgraRawImage warpedBgraRawImage = imagePerspectiveWarper.warp(bgraRawImage, corners, targetImageSize);

Document Auto Capture Controller

The DocumentAutoCaptureController interface provides a stateful document auto capture functionality.

Create DocumentAutoCaptureController:

DocumentAutoCaptureControllerConfiguration configuration = new DocumentAutoCaptureControllerConfiguration.Builder(context, validators)
    .detectionNormalizedRectangle(detectionNormalizedRectangle)
    .imageParametersNormalizedRectangle(imageParametersNormalizedRectangle)
    .onDetectionListener(this::handleDetection)
    .onCandidateSelectionStartListener(this::handleCandidateSelectionStart)
    .onCaptureListener(this::handleCapture)
    .build();
DocumentAutoCaptureController documentAutoCaptureController = DocumentAutoCaptureControllerFactory.create(configuration);

`DocumentAutoCaptureControllerConfiguration.Builder` arguments

(Required) [-] Context context – Android context.
(Required) [-] List<DocumentAutoCaptureDetectionValidator> validators – List of validators. Validations are applied in the same order in which they are stored in the list.
(Optional) [2] int minValidFramesInRowToStartCandidateSelection – Minimum number of valid frames in a row to start candidate selection.
(Optional) [1000] int candidateSelectionDurationMillis – Duration of candidate selection in milliseconds.
(Optional) [-] RectangleDouble detectionNormalizedRectangle – Crop an input image to normalized detection rectangle and use that for document detection.
(Optional) [-] RectangleDouble imageParametersNormalizedRectangle - Crop an input image to normalized image parameters rectangle and use that to analyze image parameters.
(Optional) [false] boolean mrzReadingEnabled - If Machine Readable Zone reading is enabled.
(Optional) [-] DocumentAutoCaptureController.OnDetectionListener onDetectionListener - Listener to be called on each document detection periodically.
(Optional) [-] DocumentAutoCaptureController.OnCandidateSelectionStartListener onCandidateSelectionStartListener - Listener to be called when candidate selection is started. This event happens once in a auto capture process.
(Optional) [-] DocumentAutoCaptureController.OnCaptureListener onCaptureListener - Listener to be called on auto capture process is done.

In order to get callbacks from the controller, implement these listeners from DocumentAutoCaptureController interface: OnDetectionListener, OnCandidateSelectionStartListener and OnCaptureListener.

You can use detectionNormalizedRectangle to specify the region in the input image which will be used for document detection. For example, if you want to ignore top 30% and bottom 30% of the input image, you can do it as follows:

RectangleDouble detectionNormalizedRectangle = RectangleDouble.of(0.0d, 0.3d, 1.0d, 0.7d);

If detectionNormalizedRectangle is not specified, the full input image is used for document detection.

You can use imageParametersNormalizedRectangle to specify the region in the input image which will be used to analyze image parameters. For example, if you want to ignore top 35%, left 5%, right 5% and bottom 35% of the input image, you can do it as follows:

RectangleDouble imageParametersNormalizedRectangle = RectangleDouble.of(0.05d, 0.35d, 0.95d, 0.65d);

If imageParametersNormalizedRectangle is not specified, the full input image is used to analyze image parameters.

To capture a good quality document image, repeatedly call the process() method using the camera frames:

documentAutoCaptureController.process(bgraRawImage);

The controller evaluates the document image requirements for each frame. Once there are minValidFramesInRowToStartCandidateSelection valid frames in a row, candidate selection is started with duration of candidateSelectionDurationMillis milliseconds. After the candidate selection is finished, the best document image candidate is returned by the OnCaptureListener and the document auto capture process is over.

In case you want to force the capture event, call the requestCapture() method. After you call the next process() method, the input image will be returned as a result by the OnCaptureListener and the document auto capture process will be finished.

documentAutoCaptureController.requestCapture();

In case you want to restart the document auto capture process, call the restart() method.

documentAutoCaptureController.restart();

Machine Readable Zone Reader

The MrzReader interface provides a Machine Readable Zone (MRZ) reading functionality.

Create MrzReader:

MrzReaderConfiguration configuration = new MrzReaderConfiguration.Builder(context).build();
MrzReader mrzReader = MrzReaderFactory.create(configuration);

To read a MRZ, call the following method on the background thread:

MrzReader.Result result = mrzReader.read(bgraRawImage, documentDetectorResult);

Or, alternatively if you know the travel document type call this method to order to increase precision of reading:

MrzReader.Result result = mrzReader.read(bgraRawImage, documentDetectorResult, travelDocumentType);

Argument documentDetectorResult is a product of either Document Detector component, Document Auto Capture Controller component or Document Auto Capture UI component.

The result of successful MRZ reading contains travel document type and machine readable zone. If MRZ reading was not successful, the result will contain an exception and travelDocumentType and/or machineReadableZone may be null.

UI Components

Orientation Change

In order to handle the orientation change in multi-window mode correctly, configure the activity in your AndroidManifest.xml file as follows:

<activity
    android:name=".MyActivity"
    android:configChanges="screenSize|smallestScreenSize|screenLayout|orientation" />

Document Auto Capture

The fragment with a document placeholder on the screen. The component is embedded into the application as a fragment from Android Support Library. The fragment is abstract, so it must be subclassed. In order to configure the behavior of DocumentAutoCaptureFragment, use DocumentAutoCaptureConfiguration:

DocumentAutoCaptureConfiguration documentAutoCaptureConfiguration = new DocumentAutoCaptureConfiguration.Builder().build();

Bundle arguments = new Bundle();
arguments.putSerializable(DocumentAutoCaptureFragment.CONFIGURATION, documentAutoCaptureConfiguration);

Fragment fragment = new DemoDocumentAutoCaptureFragment();
fragment.setArguments(arguments);

getSupportFragmentManager()
    .beginTransaction()
    .replace(android.R.id.content, fragment)
    .commit();

The following arguments are wrapped in DocumentAutoCaptureConfiguration:

(Optional) [CameraFacing.BACK] CameraFacing cameraFacing – Camera facing
- CameraFacing.FRONT
- CameraFacing.BACK
(Optional) [CameraPreviewScaleType.FIT] CameraPreviewScaleType cameraPreviewScaleType – The camera preview scale type
- CameraPreviewScaleType.FIT
- CameraPreviewScaleType.FILL
(Optional) [0.6d] double confidenceThreshold – Detection confidence threshold
(Optional) [0.85d] double sharpnessLowThreshold – Low sharpness threshold
(Optional) [0.25d] double brightnessLowThreshold – Low brightness threshold
(Optional) [0.9d] double brightnessHighThreshold – High brightness threshold
(Optional) [0.008d] double hotspotsScoreHighThreshold – Hotspots score threshold
(Optional) [false] boolean mrzReadingEnabled - If Machine Readable Zone reading is enabled.

As soon as the fragment is created, the camera preview starts. To start the auto capture process call the start() method. You can start the process any time.

In case you want to handle detection data, implement onDetected() callback. This callback is called with each processed camera frame. There is also onCandidateSelectionStarted() callback, which is called only once for the whole process, when candidate selection is started.

In case you want to force the capture event, call the requestCapture() method. Document image candidate will be returned via the onCaptured() callback asynchronously.

In case you want to restart the auto capture process (e.g. you want tu capture both sides of the document, one after another), call the restart() method. The whole process will start from the beginning.

Customization of UI components

Strings

You can override the string resources in your application and provide alternative strings for supported languages using the standard Android localization mechanism.

<!-- Document Auto Capture -->
<string name="dot_document_auto_capture_instruction_brightness_too_high">Less light needed</string>
<string name="dot_document_auto_capture_instruction_brightness_too_low">More light needed</string>
<string name="dot_document_auto_capture_instruction_candidate_selection">Hold still…</string>
<string name="dot_document_auto_capture_instruction_document_centering">Center document</string>
<string name="dot_document_auto_capture_instruction_document_not_present">Scan document</string>
<string name="dot_document_auto_capture_instruction_document_too_close">Move back</string>
<string name="dot_document_auto_capture_instruction_document_too_far">Move closer</string>
<string name="dot_document_auto_capture_instruction_hotspots_present">Avoid reflections</string>
<string name="dot_document_auto_capture_instruction_mrz_not_valid">Scan valid machine readable document</string>
<string name="dot_document_auto_capture_instruction_sharpness_too_low">More light needed</string>

Dimensions

You may customize the dimensions used by DOT Android Document in your application. To use custom dimensions, override the specific dimension.

<!-- Document Auto Capture -->
<dimen name="dot_document_auto_capture_placeholder_dash_path_effect_interval">8dp</dimen>
<dimen name="dot_document_auto_capture_placeholder_stroke_width">3dp</dimen>

Colors

You may customize the colors used by DOT Android Document in your application. To use custom colors, override the specific color.

<!-- Document Auto Capture -->
<color name="dot_document_auto_capture_placeholder">#ff000000</color>
<color name="dot_document_auto_capture_placeholder_capturing">#ff00ff00</color>
<color name="dot_document_auto_capture_instruction_background">#ffffffff</color>
<color name="dot_document_auto_capture_window_background">#ff000000</color>

Styles

Text views and buttons can be styled by overriding the parent style in the application.

<style name="TextAppearance.Dot.Medium" parent="TextAppearance.AppCompat.Medium" />

<style name="TextAppearance.Dot.Medium.DocumentAutoCaptureInstruction">
    ...
</style>

Common Classes

`ImageSize`

DTO which represents a size of an image. To create an instance:

ImageSize imageSize = ImageSize.of(width, height);

`BgraRawImage`

DTO which represents and an image. To create an instance:

BgraRawImage bgraRawImage = BgraRawImage.of(size, bytes);

To create an instance from Bitmap:

BgraRawImage bgraRawImage = BgraRawImageFactory.create(bitmap);

`Corners`

DTO which represents a document card corners. To create an instance:

Corners corners = Corners.of(topLeft, topRight, bottomRight, bottomLeft);

Callback DocumentAutoCaptureFragment.onDetected().
Callback DocumentAutoCaptureFragment.onCandidateSelectionStarted().

Changed

Validator threshold DocumentNotDetectedValidator.DEFAULT_THRESHOLD_CONFIDENCE to 0.6.

3.0.2 - 2021-10-08

Fixed

Restarting the Document Auto Capture UI component.

3.0.1 - 2021-10-06

Fixed

Image conversion in BgraRawImageFactory.create(Bitmap).

3.0.0 - 2021-09-27

Added

Machine Readable Zone Reader component (MrzReader, MrzReaderConfiguration and MrzReaderFactory).
Image Perspective Warper component (ImagePerspectiveWarper and ImagePerspectiveWarperFactory).
Class DocumentDetectorFactory.
Class ImageParametersAnalyzerFactory.
Class DocumentAutoCaptureControllerConfiguration.
Class DocumentAutoCaptureControllerFactory.
Class DocumentAutoCaptureDetection.
Class Corners.
Class BgraRawImageFactory.
Class BitmapFactory.

Changed

groupId com.innovatrics.android to com.innovatrics.dot.
Minimum Android API level 21.
Class DocumentAutoCaptureArguments to DocumentAutoCaptureConfiguration.
Method DocumentAutoCaptureController.detect() to DocumentAutoCaptureController.process().
Class DocumentAutoCaptureController.OnStayStillPhaseEnterListener to DocumentAutoCaptureController.OnCandidateSelectionStartListener().
Class DetectionResult to DocumentDetector.Result.
Class Point to PointDouble.
Method DocumentAutoCaptureController.onPhotoCaptured() to DocumentAutoCaptureController.onCaptured().
Method DocumentAutoCaptureController.postCaptureRequest() to DocumentAutoCaptureController.requestCapture().
Class DetectedDocument to DocumentAutoCaptureResult.
Performance of document detection is significantly improved.

Removed

Interface DocumentAutoCaptureFrameParametersEvaluator.
Class DefaultDocumentAutoCaptureFrameParametersEvaluator.
Method DocumentAutoCaptureFragment.onCameraAccessFailed().
Method DocumentAutoCaptureFragment.onCameraInitFailed().
Class PlaceholderEvaluatorFactory.
Method DocumentAutoCaptureController.onAutoCaptureReady().
Enum DocumentAutoCaptureHint.

2.3.1 - 2021-06-17

Fixed

Requesting camera permission if it is already denied.

2.3.0 - 2021-05-14

Added

DocumentAutoCaptureArguments.Builder: Method confidenceThreshold().
DocumentAutoCaptureArguments.Builder: Method sharpnessLowThreshold().
DocumentAutoCaptureArguments.Builder: Method brightnessLowThreshold().
DocumentAutoCaptureArguments.Builder: Method brightnessHighThreshold().
DocumentAutoCaptureArguments.Builder: Method hotspotsScoreHighThreshold().
BrightnessHighValidator.Builder: Method threshold().
BrightnessLowValidator.Builder: Method threshold().
DocumentDoesNotFitPlaceholderValidator.Builder: Method detectedToPlaceholderCornersDistanceThreshold().
DocumentLargeValidator.Builder: Method documentWidthToImageWidthRatioThreshold().
DocumentNotDetectedValidator.Builder: Method confidenceThreshold().
DocumentSmallValidator.Builder: Method documentWidthToImageWidthRatioThreshold().
HotspotsScoreHighValidator.Builder: Method threshold().
SharpnessLowValidator.Builder: Method threshold().

Changed

All DocumentAutoCaptureFrameParametersValidator implementations: Instance creation via a Builder.
PlaceholderEvaluatorFactory: Instance creation via a Builder.

2.2.1 - 2021-03-29

Added

Support for x86 and x86_64 architectures.

2.2.0 - 2021-03-17

Changed

Update documentation.

Fixed

Configuration change issue in Document Auto Capture UI component.

2.1.0 - 2021-01-25

Added

Callback DocumentAutoCaptureFragment.onAutoCaptureReady().
Method DocumentAutoCaptureFragment.start().

Changed

DocumentAutoCaptureFragment: Auto Capture process is not started implicitly. Client needs to call method DocumentAutoCaptureFragment.start() to start the process.
DocumentAutoCaptureFragment: As soon as the photo is captured all views are reset.

2.0.1 - 2021-01-21

Fixed

Issue in method DocumentAutoCaptureFragment.restart().

2.0.0 - 2021-01-11

Added

Possibility to restart the Document Auto Capture Controller component (method DocumentAutoCaptureController.restart()).
Possibility to restart the Document Auto Capture UI component (method DocumentAutoCaptureFragment.restart()).
New non-UI component: Image Parameters Analyzer.
DefaultDocumentAutoCaptureController.Builder: Method detectionFrame().
DefaultDocumentAutoCaptureController.Builder: Method imageParametersFrame().
HotspotsScoreHighValidator for validating hotspots presence.

Changed

Update target Android SDK version to 30 (Android 11).
DetectionResult DTO does not contain sharpness and brightness anymore (these are present in ImageParameters DTO).
DefaultDocumentAutoCaptureController.Builder: Method detectionResultEvaluator() renamed to documentAutoCaptureFrameParametersEvaluator().
DetectionResultEvaluator renamed to DocumentAutoCaptureFrameParametersEvaluator.
DefaultDetectionResultEvaluator renamed to DefaultDocumentAutoCaptureFrameParametersEvaluator.
DetectionAttributeValidator renamed to DocumentAutoCaptureFrameParametersValidator.
DetectionHint renamed to DocumentAutoCaptureHint.
Update documentation.

Removed

DocumentDetector: Method DetectionResult detectDocument(Nv21Image nv21Image).
DefaultDocumentAutoCaptureController.Builder: Method croppingFrame().

1.1.0 - 2020-11-25

Added

Possibility to force capture in Document Auto Capture Controller component (method DocumentAutoCaptureController.requestCapture()).
Possibility to force capture in Document Auto Capture UI component (method DocumentAutoCaptureFragment.postCaptureRequest()).

Changed

Update documentation.

Fixed

Proguard rules.

1.0.0 - 2020-07-31

Changed

Preview image is cropped before document detection (placeholder area and margin).
Document Auto Capture design.
Update documentation.

Fixed

Fix camera preview freezing.
Placeholder position calculation for cameraPreviewScaleType: ScaleType.CENTER_CROP.

0.3.0 - 2020-07-13

Minor changes.

0.2.0 - 2020-07-11

Changed

Update documentation.

0.1.0 - 2020-07-10

Added

First release.