Potential breakthroughs for gesture recognition with a new algorithm, a capacitive sensor


Gesture recognition technology is advancing with new software and hardware developments for digital contactless interactions.

A new hand gesture recognition algorithm has been developed by a team of researchers from Sun Yat-sen University in Guangzhou with the aim of overcoming the high computational load, low speed and precision, and limited number of recognizable gestures, reports Unite.AI.

The team, led by Zhiyi Yu, published an article describing their hand gesture recognition algorithm “for efficient advanced computing” in the Electronic Imaging Journal. The algorithm is intended to enable the deployment of gesture recognition on consumer devices, with nine gestures identifiable by the software.

Legacy algorithms can’t interpret different hand shapes, Yu explains, so the team programmed the algorithm to first classify a hand that’s presented as thin, normal, or wide. A pre-recognition step then calculates a hand area ratio to select the three most likely gestures from the nine total gestures.

“The gesture pre-recognition step not only reduces the number of calculations and hardware resources required, but also improves recognition speed without compromising accuracy,” Yu told Unite.AI.

The resulting algorithm showed real-time accuracy of up to 94%.

Innovation in capacitive sensors

Nanotechnology sensor designer Somalytics Inc. said it has developed a capacitive sensor that is the world’s smallest capable of recognizing gestures for use in digital interfaces, wellness monitoring and security applications.

The sensors are based on a 1mm paper made of carbon nanotubes, according to the announcement.

The SomaControl gesture monitor and SomaSense 3D sensing flexible floor mat are on display at the Consumer Electronics Show (CES) in Las Vegas this week. Demonstrations at CES will cover gesture control of LED lights, controlling a computer by gesture, and monitoring data such as gait and foot pressure.

Composite carbon nanotube (CPC) paper sensors are 100 times smaller and 10 times faster than existing capacitive sensors, according to the startup, which was created by CoMotion at the University of Washington with support from a investment of IP Group Inc. The sensors can detect the presence of a person at a distance of 20 centimeters.

Somalytics CEO Barbara Barclay says capacitive sensors have many potential applications in improving eye tracking.

“Somalytics’ sensors will usher in a whole new era for portable eye tracking, as the sensors are not camera-based and no eye illumination is required,” said Barclay, who according to the announcement, is a recognized international expert in eye tracking technology. “The processing speed is less than three milliseconds and the sampling rate is 10 times faster than the best available technologies. With Somalytics sensors, eye tracking will evolve to accomplish the “real-sense” and “real-time eye-to-eye” experience that augmented and virtual reality users have long been waiting for. “

Articles topics

precision | algorithms | biometrics | biometric research | advanced computing | eye tracking | walking recognition | gesture recognition | research and development | sensors


Comments are closed.