Humans have the capacity to see the world around them in detail, but image sensors for objects do not necessarily see the world the same way. They can use lower cost image sensors and a little computation.

IP licensing company Rambus Inc. (Sunnyvale, Calif.) is developing a lensless form image sensor that replaces the lens with a diffraction grating and computation. The idea is that it will provide a low-cost and non-bulky form of imaging that can be added to many things as a means of improving their contextual awareness, which could not justify the bulk and cost of lensed image sensors.

As stated in the Rambus Blog, according to inventor Dr. Patrick R. Gill, lensless technology enables sensors to capture information-rich images using a low-cost phase grating. The pinhead-sized sensor can capture all of the information in the visual world up to a certain resolution.

When the Rambus diffraction grating replaces the focusing lens, the output of a standard image sensor becomes a complex pattern of spirals, which would be meaningless to a human observer.

Rambus has developed algorithms that allow computation of position of the image field that would have produced the spiral-based pattern thereby providing a sense of vision to the object.

While an individual sensor might provide rudimentary vision that would not be acceptable to a human observer, it might be more than enough to provide a robot or airborne drone with enough visual capability to avoid obstacles.

The use of multiple lensless sensors distributed across the periphery of an autonomous object could be used to build up a sophisticated 3D map of the world around an object.

Lensless smart sensors (LSS) may be used in diverse applications and scenarios, such as helping tiny satellites orient themselves, according to Gill. Another application of would be the pairing of sensors with mini radios. Stanford researchers recently designed a prototype unit for Internet of Things (IoT) applications.

Gill also sees the use of lensless sensor technology in heavy industrial equipment such as robotics.

"A central, high-quality camera designed to triangulate the location of every possible obstacle and object of interest presents somewhat of a computational challenge for engineers," he said over at the Rambus Blog. "In contrast, packing a robotic arm with LSSs creates an effective proximity sensor that facilitates adherence to very simple rules such as basic obstacle avoidance, reaching behaviour and goal-seeking."

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved