Since the emergence of computer systems, researchers have developed various solutions in gesture recognition in the context of human-machine interaction (HMI). Nowadays, the avant-garde in this field is touchless gesture recognition, as the name indicates a sort of communication that doesn’t imply touching any sort of hardware.
Touchless technology has been incorporated in applications ranging from the industrial environment (e.g. human-robot interaction) to entertainment applications, to healthcare, and the automotive industry.
Until recently, much of the touchless gesture recognition has been focused on using computer vision techniques. However, these techniques require high computational power to understand images successfully and proved to be not helpful in the dark or other low visibility conditions. Based on the Gesture Recognition and Touchless Sensing Market Global Analysis, “sensor-based technology is anticipated to endure the largest share of the market during the next period” .
Two types of sensors are majorly utilized in touchless sensing devices: infrared sensors and capacitive sensors. However, each of these sensors has proven to have significant drawbacks. In this thesis, a design of a touchless sensor prototype has been proposed using Time-of-Flight (TOF) technology. The devised strategy aims to obtain the best configuration with a low cost and high reliability.
The proposed system is composed of a horizontal array of three TOF sensors. The sensors are employed for data collection regarding seven hand gestures, namely up, down, left, right, clockwise (CW), counterclockwise (CCW), and unknown, performed by the user at a predefined distance from the prototype. The hand gesture movements are classified using Deep Neural Networks (DNNs). Finally, the proposed approach is validated with a Graphical User Interface (GUI) representing a virtual cluster.
Moreover, this thesis offers an insight into the different technologies used in touchless sensing by providing a comparative performance evaluation of the well-known types of gesture recognition sensors available in the market.
To analyze and implement a low-cost gesture recognition system using ToF (Time-of-Flight) sensors.
Development of unconventional and low-cost human-machine interaction systems for use in specific environments.
Application of the system to different use-cases to analyze and verify the limits of the system. Use of model-based design to analyze sensor behavior, develop gesture recognition algorithms and simulate user interaction.