Neuromorphic Sensor Recognizes Motion, Predicts Path

In a single frame from a movie, a new bio-inspired smart neuromorphic sensor can identify moving objects and accurately anticipate where they will go next. Various industries, including dynamic visual sensing, automatic inspection, industrial process management, robotic guiding, and autonomous driving technology, will benefit from using this smart sensor.

Current motion detection systems could be more efficient and resource-intensive because they require several parts and intricate algorithms for frame-by-frame assessments. Researchers have created a novel neuromorphic vision technology that combines sensors, memory, and processing in a single device that can recognize motion and forecast trajectories. The human visual system inspired this technology.

An assortment of photomemristors, electrical components that generate an electric current in response to light, form the basis of their technology. When the light is turned off, the current does not end instantly. Instead, it degrades gradually, allowing photomemristors to remember if they have recently been exposed to light. As a result, unlike a camera, a smart neuromorphic sensor comprising a collection of photomemristors incorporates a dynamic memory of earlier instants and records information about the current scene.

The ability of this technology to combine several optical images into a single frame, according to the researcher, makes it special. 

Read more

Related Content: Optical Lace Enhances Soft Robot Senses