Shallow Depth of Field, Low Lateral Resolution, and Low Aspect Ratio

Micro-objective lenses’ optical lateral resolution inherently has a diffraction limit. The use of optical microscopes including digital holography has thus been limited to industries where only several micrometres of lateral resolution was required. Specifically, applications requiring 3D measurement with high aspect ratio for real time inspection of micro- or nanoscale surfaces need a high depth of field at ultra-high lateral resolution.

Numerous research groups have proposed scanning the sample in height measurement axis with a different focus plane for vertical measurement and capturing multiple holograms at a single focus plane for increasing lateral resolution. Even with big strides in the advancement of hardware, real-time processing of hundreds of captured holograms is yet out of reach.

The solution to addressing these issues lies in capturing a single but high-quality and high optical lateral resolution hologram of extended depth of field. Since this solution has yet to exist, the use and application of digital holographic microscopy has been limited to labs where measurement speed is not an issue, or to industries where high lateral resolution and height are not a requirement.

After years of extensive research and development, we are proud to present a solution technology.

Immunity to Harsh Environments

Optical holography is inherently sensitive to the slightest changes in the optical wave phase caused by airflow, acoustic waves, apparatus vibrations, and more. This sensitivity may be useful when, for example, studying micro-displacements or micro-deformations of objects under stress. However, it imposes critical limitations for holographic 3D measurements in technological environments.

Consider two types of vibrations: fast and slow.  We refer to vibrations as “fast” if their duration is much shorter than hologram capture time. The fast vibrations “smear” hologram interference patterns, reducing hologram quality. In extreme cases, the hologram cannot be recorded at all. We would therefore reduce hologram capture time so that fast vibrations will become slow ones. We call vibrations “slow” if their duration is much slower than exposure time. In this case, the vibration effect can be approximated by continuous shift of wave’s phase during the exposure.

Nonetheless, slow vibrations cause another problem that conventional holography would struggle with: they create unknown phase shifts. This becomes a significant issue for on-axis multiple phase-step holography where multiple holograms are taken with predefined phase-shifts between reference and signal waves.

When unknown phase shifts are added to predefined ones, this results in “coupling” between amplitude and phase signals, resulting in consequential errors in both reconstructed phase and depth results. This unknown phase shift can be resolved by an off-axis holography approach, but the downside is an inevitable loss of lateral image resolution.

Angstrom Vision has developed comprehensive methods to address unknown phase-shifts and holds the key to these issues.

Quick Scanning Time

In most optical 3D scanning methods, a copious number of re-focused images need to be taken. Confocal microscopy and white light interferometry are examples of such approaches. For these methods, it is not uncommon to have an image per vertical resolution element. A lengthy scanning process of capturing 128 images will result in constructing a 3D map with only 8-bit depth resolution.

Coherent holography, on the other hand, provides exceptionally high-resolution depth maps at a quick scanning pace. For instance, we can create 10-bit depth maps at the frame rate of the digital camera being used. Despite an extremely high height measurement range in 3D space (e.g., 10mm), our invention only requires a single image to reconstruct a 3D depth map. This method drastically reduces system processing time and can thus be installed where fast-moving targets need precise measurement.

Until recent years, this advantage of coherent digital holography had been overweighed by its disadvantage: huge speckle-noise. Angstrom Vision has invented new optical system structure and image-processing algorithms that are proven to be highly efficient for speckle-noise suppression.

Advanced Image Processing

The main issue of quantitative holographic measurements is the speckle noise. This noise had been studied for decades and efficient methods have recently emerged for suppressing it in amplitude signals.

For holographic 3D shape measurement, phase signal is of main interest–we therefore developed efficient speckle noise suppressing algorithms for phase signals.

The effect of speckle-noise suppression is illustrated in the figure below. It shows 3D surface screening results by a conventional holographic two-wave method on the left, and the Angstrom Vision approach on the right. The conventional method demonstrates heavy speckle noise that hinders 3D measurement accuracy; with the advanced Angstrom Vision method, the 3D shape measurement accuracy is at a sub-micron level.

It is also worth noting that our advanced image processing algorithms are implemented in a computationally efficient (i.e., fast) manner. Consequently, when being run on a conventional GPU or FPGA, it achieves real-time processing at video frame rates for 24MP depth maps.

Small Form Factor

One of our most recent innovations is enhancing the beam-steering technique, which enables miniaturization of the holographic part of device design and offers compact holographic add-ons to existing photographic inspection systems and hand-held 3D scanners.

This innovation is based on the fusion of holography and integrated optics and would theoretically help build compact systems with unparalleled 3D scanning performance in both accuracy and speed.

Additional benefits include optimizing the size and weight of large hardware, simplifying the number of parts, quick and smooth movement of the Microsoft header, and ability to install multiple heads for swift measurement of a large region of interest (e.g., a large display panel, glass, or metal).