When it comes to smartphone cameras, bigger is better. Larger image sensors and lenses are able to process more light and therefore resolve more detail. This is especially important because the filters that create color images also block about 70% of incoming light.

These color filters - which come as a grid of red, green, and blue over the pixels of an image sensor - have been around for decades. But new approaches promise to exploit the physics of light to create color images without blocking too many photons. Three approaches to achieving sharper images were presented at the 2023 IEEE International Electron Devices Meeting (IEDM). Now, these approaches are starting to move beyond the lab stage.

Samsung, for example, will supply the front-facing camera for a new phone from China's Xiaomi that uses Samsung's nanoprism technology to improve low-light performance. Rather than replacing color filters, the technology uses the principle of diffraction to collect more light in each pixel of a specific color. According to the company, this can increase sensitivity by 25%.

Meanwhile, two startups have developed ways to capture color images without filters. Imec spinoff Eyeo announced this month that it had raised €15 million in seed funding. PxE Holographic Imaging demonstrated technology that combines depth sensing and color imaging at the Consumer Electronics Show (CES) in January.

Both PxE and Eyeo are compatible with CMOS sensors, the most common digital image sensors used in today's cameras. "CMOS sensors are a very mature and powerful platform that is worth building on. Every device has them today," said Yoav Berlatzky, founder and CEO of PxE. But "everyone wants more photons to reach their CMOS sensors."

Eyeo's filter-free color camera

Eyeo aims to commercialize the research results presented by Imec at IEDM 2023 for applications in consumer electronics, security and other fields. By removing the color filter, the startup's image sensor will be three times more sensitive than traditional CMOS sensors. "It's like we have finally opened the eyes of the image sensor," said Jereon Hoet, CEO of Eyeo.

The color separators in Eyeo image sensors direct the different wavelengths of light to the appropriate pixels.

It works by feeding light into a vertical waveguide, which splits the light according to wavelength and then directs the photons to the appropriate pixels. The waveguide acts like a funnel, so the width of these pixels can be reduced to less than 0.5 microns, about half the size of a typical smartphone pixel. According to Imec's research, the technology is also more consistent with the color sensitivity of the human eye than current filter-based imagers.

The color separation technology is designed to be implemented using existing tools and processes at CMOS foundries. The challenge lies in the software. According to Hoet, Eyeo is currently working to ensure that the sensor is compatible with the systems of its potential customers.

In terms of applications, Hoet said that the advantages of Eyeo's smaller and more sensitive image sensors are particularly evident in the field of smartphones. However, he expects that the technology will first be used in other fields, such as security systems in low-light conditions or augmented reality devices that require ultra-compact sensors.

PxE brings 3D technology to CMOS

The basic ideas behind the PxE approach are similar. Both companies are working to simulate color filters without losing photons and to bend light waves "in a way that puts the color in the right place on the right pixel," Bertlatzky concluded.

In the photo above, red lines indicate objects that are closer, while blue lines indicate objects that are farther away.

PxE's technology uses a layer of diffractive material called a "holographic encoder" that not only creates a color image but also acts as a depth sensor (hence the word "holographic" in the company's name). When white light passes through the holographic encoder, it creates an interference pattern that is recorded by the sensor. PxE's algorithm then uses this interference pattern to reconstruct a virtual 3D image - a hologram. The interference pattern also encodes information about the wavelength of light, so a color (and infrared) image can be reconstructed at the same time. Berlatzky says PxE's hardware is "cleaner" than color splitters and other approaches that use specially designed metasurfaces. Most of its power comes from software. "The algorithm is based on the physics of light," Berlatzky explains. "You can think of it as us running backwards, from the CMOS sensor back to the real world, and reconstructing the depth of field and image that the camera actually saw. Like Eyeo, PxE's image sensors can be used in a variety of applications, especially those that already have standalone depth and image sensors, such as automotive and smartphones.

Reference link https://spectrum.ieee.org/cmos-color-filter

Source: Content from IEEE Spectrum

View more at EASELINK