How-to: Install and Use LumiScan Annotation Framework for Oriented Bounding Box Annotation

Introduction

This how-to guide describes how to install and use the LumiScan Annotation Framework for creating oriented bounding box annotations.

The document explains how to:

• Install the LumiScan Annotation Framework on Windows

• Configure the application

• Load image datasets

• Create and modify oriented bounding boxes

• Save annotation data

This HOW-TO focuses on software installation and operation only.

Detailed recommendations and best practices for creating high-quality annotations are described in a separate annotation guideline HOW-TO.

The LumiScan Annotation Framework is used to create annotation data for training and retraining neural network models used in applications such as:

• Object detection

• AI Object Matching

• Quality inspection

Prerequisites

Used Hardware

• Windows PC

• Mouse and keyboard

• Optional:

  • CmDongle license device

Used Software

• LumiScan Annotation Framework installer

• CodeMeter Runtime Environment

Operating System Requirements

• Windows 10 or higher

System Overview

The LumiScan Annotation Framework allows manual annotation of images using oriented bounding boxes.

The software supports:

• Multiple users

• Multiple image folders

• Unlimited annotations

• Rotated bounding boxes

Annotation data is automatically stored together with the images.

Step-by-Step Guide

1. Verify CodeMeter Installation

The LumiScan Annotation Framework requires the CodeMeter Runtime Environment for licensing.

Before installing the application:

1. Verify whether CodeMeter Runtime is already installed.

2. If CodeMeter is not installed:

   • Download the installer from:

https://www.wibu.com/de/support/anwendersoftware/anwendersoftware.html

1. Install CodeMeter Runtime version 7.60c or higher.

2. Connect the CmDongle license device to a free USB port if required.

2. Install LumiScan Annotation Framework

1. Close all running applications.

2. Navigate to the LumiScan Annotation Framework installer location.

3. Start the installer:

Installer_LumiScan_Annotation_va_b_c_d.exe

1. Accept the license agreement.

2. Select the installation directory if required.

3. Continue the installation process.

4. Click:

   • Install

   • Finish

After installation, the application can be started via the Windows Start Menu or desktop shortcut.

3. Prepare the Dataset Structure

To annotate images, the dataset must contain PNG images.

For a single dataset:

• Place all .png files inside one folder

For multiple datasets, the following structure is recommended:

<Dataset Root>
|-----<Folder A>
|     |---- image_001.png
|     |---- image_002.png
|
|-----<Folder B>
      |---- image_003.png
      |---- image_004.png

The folder selection inside the application can then be used to switch between datasets.

4. Start the Application

1. Start LumiScan Annotation Framework.

2. When started for the first time, an empty application window appears.

3. Open:

   Menu → Settings

4. Configure the data directory.

5. Click:

   • Set Folder

6. Select the dataset directory containing the images.

After selecting the dataset:

• The images appear in the image list

• The folders become selectable via the folder selection field

5. Review the User Interface

The main user interface contains:

• Menu access

• Folder selection

• Image list

• Label list

• Central image view

• Navigation buttons

The central image view displays:

• The currently selected image

• Existing annotations

• ROI borders

Brightness and contrast can be adjusted using the sliders on the right side.

6. Create Labels

Labels define the object classes used during annotation.

To create a new label:

1. Enter the label name in:

   • Add Label

2. Click:

   • Add Label

The label then appears in the label list and can be assigned to bounding boxes.

Fig. 3: Label creation The screenshot should show the label list and the Add Label input field.

7. Create Oriented Bounding Boxes

The LumiScan Annotation Framework uses oriented bounding boxes for annotation.

To create a new bounding box:

1. Double-click with the left mouse button at the desired bottom edge position.

2. Click the upper edge of the object to define the height.

3. Click a side point of the object to define the width.

The annotation is automatically assigned to the currently selected label.

8. Modify Existing Bounding Boxes

Existing bounding boxes can be adjusted after creation.

To select a bounding box:

• Left-click the bounding box

Selected boxes are highlighted and display corner handles.

The selected box can then be:

• Moved

• Resized

• Rotated

• Relabeled

To delete a bounding box:

• Right-click the selected box

9. Save Annotations

To save annotations:

1. Click:

   • Save

or: 2. Press:

• s

The application stores:

• Annotation CSV files

• Metadata JSON files

The generated JSON metadata includes:

• User name

• Save timestamp

• Image reference

• Save file name

10. Navigate Through Images

To switch between images:

• Click:

  • Previous

  • Next

The navigation buttons are located at the bottom of the application window.

The application also supports copying annotations to the next image using:

• Save & Copy

This is useful when image content changes only slightly between consecutive images.

Result

After completing the steps described in this guide:

• LumiScan Annotation Framework is installed successfully

• Image datasets can be loaded

• Labels can be created

• Oriented bounding boxes can be added and modified

• Annotation data is stored correctly for neural network training

For annotation recommendations and labeling quality guidelines, refer to the dedicated annotation guideline HOW-TO.

The Company

HD Vision Systems is a leading provider of industrial image processing and AI-based automation solutions. Our hardware-agnostic platform enables seamless integration across a wide range of cameras, sensors, robots, and industrial systems, giving customers maximum flexibility and future-proof scalability. By combining advanced machine vision software with strong partnerships across the automation ecosystem, we deliver tailored, high-performance solutions for quality inspection, process optimization, and intelligent manufacturing. Together with our technology partners, we help companies accelerate digital transformation and unlock the full potential of Industry 4.0.