Augmented reality adds digital content onto a live camera feed (or translucent screen), making that digital content look as if it is part of the physical world around you. To accurately place the content in the real-world environment, your device has to get a reference point. 

Some applications use the so-called "markers" as a reference, i.e., any image complex enough to be recognized as distinct and put the content on top. Other applications use more advanced algorithms like ARKit or ARCore, allowing us to place the content on any flat surface with a distinctive texture. 

AR based on markers vs. markerless AR

Marker-based AR is the simplest form of augmented reality using image recognition. Our device's camera recognizes a predefined image (marker) and places the digital content on top of it. Almost any distinctive image can be used as a marker, from a printed QR code to your company's logo or a magazine page. We can display various types of content on top of the marker, such as videos, animations, 3D models, and much more. We can also add multiple types of interactions to the experience to make it more exciting and engaging. Marker-based AR can be run on almost any kind of smartphone with a working gyroscope and a camera.

Markerless AR - a more advanced form of augmented reality, it uses an algorithm called SLAM (simultaneous localization and mapping) to recognize flat surfaces and allow us to place the content on top of the chosen surface, without the need for a marker. The most popular AR engines for markerless AR are ARKit (for Apple devices) and ARCore (for Android devices). Both engines are continually being developed, introducing new and exciting functionalities like light estimation, human occlusion, and many more.