Whether it be drones, mobile robots or self-driving cars, it’s incredibly important that robots have good vision to see their surroundings. There have been major improvements in this area over the years, of course, and now researchers at the University of Stuttgart have developed a tiny 3D-printed camera lens that allows for speedier image processing.
The camera uses four lenses instead of one, each set at different focal lengths and mounted on an image-reading microchip that compiles data from all four lenses into a single image. Each lens is made from plastic and is no bigger than a grain of salt.
But the true magic here is that the micro-lenses mimic something called foveated vision – read about this in-depth over at Science Advances – which eagle and human eyes have to create a wide field of view at low resolution and focus on a single object at high resolution at the same time. The researchers working on the micro-lenses say drones, sensors for self-driving cars, and vision systems for robots are just some of the applications that could benefit from a higher resolution at the center of their field of view.
The researchers do admit, however, that the lenses have some limitations. For example, the resolution is low, and it takes several hours to 3D print each lens. The researchers are quite confident that these kinks will be worked out going forward.
Comparison of simulation and measurement of foveated imaging systems.
“By using more advanced image fusion algorithms, it will be possible to further improve the image quality and increase the resolution toward the center in a smoothly varying fashion. However, the implementation of these approaches is beyond the scope of this paper,” the researchers write.
Here’s more from the researchers:
“Our work demonstrates for the first time direct 3D printing of varying complex multicomponent imaging systems onto a chip to form a multiaperture camera. We combine four different air-spaced doublet lenses to obtain a foveated imaging system with an FOV of 70° and angular resolutions of >2 cycles/deg in the center of the image. At the moment, the chip dimensions and pixel size limit the overall systems dimensions and our optical performance, whereas future devices can become smaller than 300 μm × 300 μm × 200 μm in volume and, at the same time, transfer images with higher resolution. The method thus enables considerably smaller imaging systems as compared to the state of the art.”
Working principle of the 3D-printed foveated imaging system. (Credit: Science Advances)