Researchers revive the pinhole camera for next-gen infrared imaging

Stephanie Baum
scientific editor

Robert Egan
associate editor

Researchers have used the centuries-old idea of pinhole imaging to create a high-performance mid-infrared imaging system without lenses. The new camera can capture extremely clear pictures over a large range of distances and in low light, making it useful for situations that are challenging for traditional cameras.
"Many useful signals are in the mid-infrared, such as heat and molecular fingerprints, but cameras working at these wavelengths are often noisy, expensive or require cooling," said research team leader Heping Zeng from East China Normal University. "Moreover, traditional lens-based setups have a limited depth of field and need careful design to minimize optical distortions. We developed a high-sensitivity, lens-free approach that delivers a much larger depth of field and field of view than other systems."
Writing in Optica, the researchers how they use light to form a tiny "optical pinhole" inside a nonlinear crystal, which also turns the infrared image into a visible one. Using this setup, they acquired clear mid-infrared images with a depth of field of over 35 cm and a field of view of more than 6 cm. They were also able to use the system to acquire 3D images.
"This approach can enhance night-time safety, industrial quality control and environmental monitoring," said research team member Kun Huang from East China Normal University. "And because it uses simpler optics and standard silicon sensors, it could eventually make infrared imaging systems more affordable, portable and energy efficient. It can even be applied with other spectral bands such as the far-infrared or terahertz wavelengths, where lenses are hard to make or perform poorly."
Pinhole imaging reimagined
Pinhole imaging is one of the oldest image-making methods, first described by the Chinese philosopher Mozi in the 4th century BC. A traditional pinhole camera works by letting light pass through a tiny hole in a lightproof box, projecting an inverted image of the outside scene onto the opposite surface inside. Unlike lens-based imaging, pinhole imaging avoids distortion, has an infinite depth of field and works across a wide range of wavelengths.
To bring these advantages to a modern infrared imaging system, the researchers used an intense laser to form an optical hole, or artificial aperture, inside a nonlinear crystal. Because of its special optical properties, the crystal converts the infrared image into visible light, so that a standard silicon camera can record it.
The researchers say that the use of a specially designed crystal with a chirped-period structure, which can accept light rays from a broad range of directions, was key to achieving a large field of view. Also, the upconversion detection method naturally suppresses noise, which allows it to work even in very low light conditions.
"Lensless nonlinear pinhole imaging is a practical way to achieve distortion-free, large-depth, wide-field-of-view mid-infrared imaging with high sensitivity," said Huang. "The ultrashort synchronized laser pulses also provide a built-in ultrafast optical time gate that can be used for sensitive, time-of-flight depth imaging, even with very few photons."
After figuring out that an optical pinhole radius of about 0.20 mm produced sharp, well-defined details, the researchers used this aperture size to image targets that were 11 cm, 15 cm and 19 cm away. They achieved sharp imaging at the mid-infrared wavelength of 3.07 μm, across all the distances, confirming a large depth range. They were also able to keep images sharp for objects placed up to 35 cm away, demonstrating a large depth of field.
3D imaging without lenses
The investigators then used their setup for two types of 3D imaging. For 3D time-of-flight imaging, they imaged a matte ceramic rabbit by using synchronized ultrafast pulses as an optical gate and were able to reconstruct the 3D shape with micron-level axial precision. Even when the input was reduced to about 1.5 photons per pulse—simulating very low-light conditions—the method still produced 3D images after correlation-based denoising.
They also performed two-snapshot depth imaging by taking two pictures of a stacked "ECNU" target at slightly different object distances and using those to calculate the true sizes and depths. With this method, they were able to measure the depth of the objects over a range of about 6 centimeters, without using complex pulsed timing techniques.
The researchers note that the mid-infrared nonlinear pinhole imaging system is still a proof-of-concept that requires a relatively complex and bulky laser setup. However, as new nonlinear materials and integrated light sources are developed, the technology should become far more compact and easier to deploy.
They are now working to make the system faster, more sensitive and adaptable to different imaging scenarios. Their plans include boosting conversion efficiency, adding dynamic control to reshape the optical pinhole for different scenes, and extending the camera's operation across a wider mid-infrared range.
More information: Yanan Li et al, Mid-infrared nonlinear pinhole imaging, Optica (2025).
Journal information: Optica
Provided by Optica