✨ Why iMarkers?

Invisible Landmarks for Robots

iMarkers are discreet landmarks for robotics and AR applications, invisible or barely noticeable to the naked eyes when placed in an environment.

Flexible and Versatile Production

iMarkers can be produced to be detected in particular light wavelengths (UV, IR, or visible), enabling easy customization for diverse applications.

Cost-efficient Fabrication and Detection

iMarkers fabrication is rapid and cost-effective, with detection achievable using affordable, custom-built sensors, ensuring high accessibility.

Fast and Robust Recognition

iMarkers can be detected and recognized using low computation cost and robust computer vision algorithms.

📸 Media

🔦 iMarker Variants

iMarkers offer production flexibility, enabling design and detection across the entire light spectrum. These variants include UV-range (entirely transparent and invisible to the naked eye), IR-range (similarly discreet), and visible-range iMarkers designed to blend with patterned backgrounds or camouflage against similarly colored surfaces.

  • A 6cm × 6cm IR-range iMarker, fabricated with ArUco patterns
  • A 6cm × 6cm UV-range iMarker revealed after UV illumination
  • A 6cm × 6cm UV-range iMarker placed on a transparent surface
  • A 6cm × 6cm ArUco-patterned IR-range iMarker placed on a textured background
  • A 7cm × 7cm visible-range iMarker (in hand) and a printed 10cm × 10cm ArUco markerr in the background

📷 iMarker Detector Sensors

Due to the versatility of iMarkers, various sensors can be designed and employed to detect and recognize them. So far, three primary solutions are proposed, including “dual-vision,” “dynamic single-vision,” and “static single-vision.” You can see some implemented setups below.

  • Dual-vision setup, containing to cameras facing a cube beamsplitter
  • Dual-vision setup, containing to cameras facing a plate beamsplitter
  • Single-vision setup with a fixed plarizer attached to the camera's RGB lens
  • Single-vision setup with a circular polarizer switching between left- and right-handed modes, driven by a function generator

💻 iMarkers Detector Algorithms

Regarding various sensors and iMarker types, different algorithms are proposed to detect and recognize iMarkers. These algorithms are designed to be fast, robust, and computationally efficient, ensuring real-time performance. Below, you can see some of the implemented algorithms for different iMarker types and sensor setups.

  • Detecting a visible-range iMarker using the static single-vision sensor
  • Detecting a visible-range iMarker using the dual-vision setup
  • Detecting a visible-range iMarker using dynamic single-vision setup
  • Detecting a UV-range iMarker using the single-vision setup (UV flashlight used)

Highlights

⚠️ Content has been anonymized as the paper is currently undergoing double-blind peer review.

Team

⚠️ Content has been anonymized as the paper is currently undergoing double-blind peer review.

Publications

⚠️ Content has been anonymized as the paper is currently undergoing double-blind peer review.

Contact Us

⚠️ Content has been anonymized as the paper is currently undergoing double-blind peer review.