Smartphone frames an indoor environment for visual positioning
GiPStech for developers

AIDK: AI visual positioning

Use camera images to help your app understand where a person is and which floor they are on. AIDK adds a visual estimate to the location flow when a space calls for more context than movement and radio signals alone can provide.

01 The component

When the environment becomes a reference

Inside a building, images can provide clues to location. AIDK applies AI models to camera frames to estimate a place and, when the relevant model is available, the floor. It is separate from MDK, so you can add visual positioning when your project needs it. The AIDK overview explains its role.

Developer reviews an indoor visual-location flow

Images become another reference for location.

02 Capabilities

An estimate that complements location

The image-based position can be combined with other signals handled by MDK. Your app can use visual data as another reference in the location flow, especially where recognizing the surrounding context matters.

Results also depend on the images used to prepare the model and on real-world conditions. The visual positioning guide covers capabilities and requirements.

Multi-level indoor environment with visual landmarks
03 The spaces

Prepared for your spaces

Models are organized for the spaces and levels they need to recognize. Automatic floor estimation is available when the corresponding model has been prepared; otherwise the app can manage the active level. It helps to define buildings, floors and survey plans early, then validate the solution on site.

04 Integration

An added capability within a coherent flow

ARDK provides camera access, AIDK interprets the image and MDK can combine the estimate with other location signals. Each component has a distinct role, and they can be brought together around the experience your app needs. The technical guide explains the connections.

AIDK

Interprets images with AI models.

ARDK

Provides camera access.

MDK

Combines the estimate with other location signals.

05 The flow

From image to experience

The app flow has three stages: capture an image, estimate the location with a model prepared for the space and use the result in the location service. The technical guide explains how to connect the components.

01

Capture

Collect an image in the app.

02

Estimate

Use a model prepared for the space.

03

Integrate

Bring the result into the location service.

06 Adoption

From trial to product

We can review devices, image quality and results in representative spaces before extending the service. Once activated and with models available on the device, AIDK can operate offline; activation and licence renewal require a connection. For distribution and requirements, see the installation documentation.

Developer prepares a mobile project in a work environment

Explore visual positioning for your project

Tell us about your spaces and the service you want to offer. We can define how to prepare and test the models.

Wait a second…

Stay in the loop with our latest updates.

Subscribe to our newsletter for exclusive insights, product updates, and more.

We respect your inbox – no spam, ever.

GDPR Consent Form