Reader SDK Documents
Breadcrumbs

Novarum Reader SDK Android Integration Guide

1.1. Overview

The Novarum SDK enables Android applications to capture and analyze lateral flow tests using the device’s camera.
It handles camera setup, image analysis, and result delivery, allowing developers to focus on interpreting and presenting test results in their own UI.

This guide describes how to integrate the SDK into an Android project using Gradle and AWS CodeArtifact.


1.2. Repository Access

1.2.1. 1 AWS CLI Configuration

Access to the Novarum SDK Maven repository is provided via AWS CodeArtifact.
Before adding the dependency, you must install and configure the AWS CLI.

1.2.1.1. Installation

Refer to AWS Command Line Interface for installation details.
On macOS, you can install via Homebrew :

brew install awscli

1.2.1.2. Profile Configuration

You’ll need the AWS credentials provided by Novarum.

Interactive setup:

aws configure --profile ndxcodeartifact

You’ll be prompted for:

  • Access Key ID

  • Secret Access Key

  • AWS Region (e.g., eu-west-2)

For non-interactive environments (CI, build agents), configure directly:

aws configure set aws_access_key_id "${ECR_KEY}"
aws configure set aws_secret_access_key "${ECR_SECRET}"
aws configure set region "${AWS_REGION}"

1.3. Project Setup

1.3.1. 1 Minimum Requirements

Requirement

Version

Minimum SDK

26

compileSdk

35

Java / JVM Target

17

Android Gradle Plugin

8.9.0

Gradle Wrapper

8.11.1

Kotlin

1.9+


1.3.2. 2 Maven Repository Configuration

Add the Novarum Maven repository to your root build.gradle or settings.gradle file.

This configuration retrieves an authorization token for each build from AWS CodeArtifact.

def codeartifactToken = "aws codeartifact get-authorization-token --domain novarumdx --domain-owner 945969778369 --query authorizationToken --output text --profile ndxcodeartifact".execute().text.trim()

allprojects {
    repositories {
        maven {
            credentials {
                username = 'aws'
                password = codeartifactToken
            }
            url 'https://novarumdx-945969778369.d.codeartifact.eu-west-2.amazonaws.com/maven/novarum.client.maven/'
        }
        google()
        mavenCentral()
    }
}

1.3.3. 3 Gradle Configuration

Ensure your module build.gradle uses the following settings:

android {
    namespace "com.example.app"
    compileSdk = 35

    defaultConfig {
        minSdk = 26
        targetSdk = 35
    }

    compileOptions {
        sourceCompatibility JavaVersion.VERSION_17
        targetCompatibility JavaVersion.VERSION_17
    }

    kotlinOptions {
        jvmTarget = "17"
    }
}

dependencies {
    implementation "org.novarumdx:camerax:<version>"
}


1.4. Android Manifest Configuration

The SDK requires access to the camera.
Add the following entries to your app’s AndroidManifest.xml:

<uses-permission android:name="android.permission.CAMERA"/>
<uses-feature android:name="android.hardware.camera" android:required="true" />

If your app targets Android 13+ and uses runtime permissions, request android.permission.CAMERA before starting a scan.

Request camera permission before displaying NovarumAnalyzerPreview.


1.5. Using the Reader View

The Novarum Analyzer Preview is the primary camera component used to display the live feed and perform the test analysis.
It can be embedded in either a traditional XML layout or a Jetpack Compose UI, depending on your project setup.


1.5.1. XML-Based Integration

For standard View-based layouts, include the NovarumAnalyzerPreview directly in your XML layout file:

<org.novarumdx.camerax.NovarumAnalyzerPreview
    android:id="@+id/previewView"
    android:layout_width="match_parent"
    android:layout_height="match_parent" />

Then configure it in your Activity or Fragment:

val configJson = /* your analyser configuration JSON */
val analyserConfiguration = PMFLoader.loadAnalyserConfiguration(json = configJson)

binding.previewView.analyzerConfig = analyserConfiguration
binding.previewView.previewMode = PreviewMode.Fill

binding.previewView.onFrameCaptured = { callback ->
    // Track scan progress, brightness, orientation, etc.
    val progress = callback.progress
}

binding.previewView.onComplete = { result ->
    // Handle analysis completion
}

val result = binding.previewView.abort() // Aborts the scan if required

1.5.2. Compose Integration

The SDK also provides a ReaderView composable for use in Jetpack Compose projects.
It wraps NovarumAnalyzerPreview internally and provides the same functionality with Compose-friendly bindings.

@Composable
fun AnalyzerScreen(analyserConfiguration: AnalyserConfiguration, onComplete: (ResultModel) -> Unit) {
    ReaderView(
        analyserConfiguration = analyserConfiguration,
        previewMode = PreviewMode.Fill,  // or PreviewMode.Fit
        onComplete = onComplete,
        onAbort = { result ->
            // Handle abort event (result may be null if no frames processed)
        }
    )
}

1.5.2.1. Example With UI Controls

@Composable
fun AnalyzerScreen(analyserConfiguration: AnalyserConfiguration) {
    var progress by remember { mutableFloatStateOf(0f) }
    ReaderView(
        analyserConfiguration = analyserConfiguration,
        previewMode = PreviewMode.Fill,
        onFrameCaptured = { callback ->
            progress = callback.progress
        },
        onComplete = { result ->
            // Handle analysis result
        },
        onAbort = { result ->
            // Handle abort event
        }
    )
}

1.5.3. Parameter Overview

Parameter

Type

Description

analyserConfiguration

AnalyserConfiguration

The analyser configuration JSON object defining the test setup and analysis parameters.

onComplete

(ResultModel) -> Unit

Invoked when the analysis successfully completes. Provides the full analysis result.

onAbort

(ResultModel?) -> Unit

Called when the scan is aborted manually via preview.abort(). Returns partial results or null if none collected.

onFrameCaptured

(FrameCapturedCallback) -> Unit

Triggered for every frame processed by the reader. Useful for updating progress indicators or monitoring lighting/orientation.

previewMode

PreviewMode

Controls camera scaling within the view. Options:

  • PreviewMode.Fill – fills the container (may crop edges).

  • PreviewMode.Fit – fits entirely within the view while maintaining aspect ratio.

To abort a running scan, call preview.abort() on the NovarumAnalyzerPreview instance.
The onAbort callback will then receive the partial analysis data.


1.5.4. Callback Data Structures

The SDK exposes several structured models representing frame-level and result-level data.

1.5.4.1. FrameCapturedCallback

data class FrameCapturedCallback(
    val progress: Float,
    val resultPMF: ResultPMF,
    val stripStatuses: List<StripStatus>,
    val orientation: List<Double>,
    val lux: Float,
)

Property

Type

Description

progress

Float

Current scan progress (0.0 – 1.0).

resultPMF

ResultPMF

Frame-level fit classification (Fit, MarginalFit, NoFit).

stripStatuses

List<StripStatus>

Current status of each detected strip.

orientation

List<Double>

Device orientation at capture time.

lux

Float

Ambient light level.


1.5.4.2. ResultModel

data class ResultModel(
    val testConfiguration: TestConfiguration,
    val testStrips: List<TestStrip>,
    val pmfStory: List<FrameData>,
)

Property

Type

Description

testConfiguration

TestConfiguration

Test setup and analyser parameters.

testStrips

List<TestStrip>

List of analysed strips with control/test line data.

pmfStory

List<FrameData>

All captured frame data during the analysis session.


Type

Kind

Description

ResultPMF

enum

Indicates the per-frame fit result (Fit, MarginalFit, NoFit).

StripStatus

enum

Indicates strip condition (None, ExposureError, BaselineError, Good, Done).

TestStrip

data class

Contains data for one analysed strip (C-line, T-lines, profiles, baselines).

FrameData

data class

Metadata for each captured frame (lighting, homography, orientation, etc.).

For full type definitions, refer to the generated documentation within Android Studio.


1.5.6. Behaviour Notes

  • previewMode determines how the live camera preview scales to your layout. Fill (default) maximises coverage, while Fit ensures full frame visibility.

  • The abort() method immediately stops scanning and triggers onAbort, returning partial data.

  • Both XML and Compose integrations share the same underlying NovarumAnalyzerPreview logic.


1.6. Optional: SVG Overlay Visualisation

The SDK supports an optional SVG-based overlay rendering mode for higher-quality graphics and scalable test result visuals.

  • Default visualisation mode: dotMatrix

  • To enable SVG overlays, request an updated Analyser Configuration File from Novarum support.

  • No code changes are required — the overlay mode is controlled via configuration.


1.7. Embedded Assets

The SDK now includes dedicated image assets for exposure strip errors and other visual cues.
These assets are embedded within the SDK — no customisation or configuration is required by developers.


1.8. Native Code Considerations

If your app or modules build native (JNI/C++) components:

  • Update to NDK r29.0.14033849 rc4

  • Update CMake to 3.22.1

  • Rebuild your project with:

    ./gradlew clean assembleRelease
    

This ensures compatibility with the SDK’s OpenCV 4.12.0 dependency.


1.9. Troubleshooting & Support

Issue

Possible Cause

Resolution

Build fails with “invalid target JDK”

Using JDK 11 or 1.8

Update Gradle JDK to 17

AGP / Gradle mismatch

AGP < 8.9.0

Update Android Gradle Plugin to 8.9.0 and Gradle Wrapper to 8.11.1

Cannot access CodeArtifact

Invalid or expired AWS token

Regenerate token or verify AWS CLI profile

Overlay not scaling correctly

Outdated configuration file

Request updated analyser configuration (SVG enabled)

For further assistance or updated configuration files, contact your Novarum technical representative.


1.10. Version Compatibility Summary

Component

Current Version

Notes

Novarum SDK

v3.0.0

Current release

Minimum SDK

26

Required

compileSdk

35

Required

Java / JVM Target

17

Required

Gradle

8.11.1

Required for AGP 8.9.0

AGP

8.9.0

Required

OpenCV

4.12.0

Internal dependency

NDK

29.0.14033849 rc4

Required for native builds

CMake

3.22.1

Required for native builds