Computer Vision Interface for Symbolic Programming of Cartesian Motion to introduce Visually Impaired Children into Robotic Sciences

Didactic material, aimed at sightless children interested in Robotics, should provide a suitable working environment for their specific sensory capacities and instill a spatial perception of motion and an ability to describe it in a logical and structured way. In this paper, to improve the user expe...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International Conference on Electrical Engineering, Computing Science, and Automatic Control (Online) S. 1 - 6
Hauptverfasser: Cavazos-Carrizales, Juan P., Ruiz-Sanchez, Francisco J.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 09.11.2022
Schlagworte:
Computer vision
Robot motion
Robot sensing systems
Symbols
Three-dimensional displays
Visualization
Webcams
ISSN:2642-3766
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Didactic material, aimed at sightless children interested in Robotics, should provide a suitable working environment for their specific sensory capacities and instill a spatial perception of motion and an ability to describe it in a logical and structured way. In this paper, to improve the user experience learning the basics of coding robotic motion in a non-visual environment, we introduce a symbolic programming interface of assembling tokens based on artificial vision recognition that preserves the working space free of obstacles and simplifies the programming process by sliding tokens together on a flat board instead of requiring complex electromechanical connections between them. The computer vision system detects symbols and numbers based on a Boundary Object approach that extracts the main features of the captured images and assigns them to the closest class contained in a reference library. Detected symbols and numbers are identified as part of an ordered array and interpreted as commands with an argument, providing auditory assistance in the case of errors, before being executed in a real mobile robot for instant learning reinforcement. Tokens were ergonomically redesigned for children allowing a simple manipulation for easy assembling and providing tactile and visual information to allow easy recognition by users, their instructors, and the artificial vision system. Throughout the paper, we discuss the design criteria before presenting the final implementation and its first results.
ISSN:2642-3766
DOI:10.1109/CCE56709.2022.9975943