Traditional microscopes and 3D imaging setups have long been constrained by the inflexibility of their recorded images. Once a hologram or microscopic image is captured, conventional methods do not allow alterations to key imaging parameters such as focal depth, limiting the ability to optimize or tailor images for detailed analysis later. Addressing this challenge, Shivasubramanian Gopinath, a Junior Research Fellow at the University of Tartu, alongside his colleagues, has pioneered a method that captures multiple holograms at varying focal distances simultaneously at the time of acquisition.

Unlike classical single-shot holography, this technique enables the acquisition of a set of holographic data representing different focal planes. These multiple recordings are then computationally combined using an advanced reconstruction algorithm, creating a synthetic hologram endowed with dramatically improved depth of focus. This computational post-processing approach transforms the rigid nature of holograms, allowing scientists to adjust imaging properties after capture, tailoring images to the needs of precise scientific analysis.

This breakthrough builds on the foundation of Fresnel Incoherent Correlation Holography (FINCH), a well-established method for recording three-dimensional information under incoherent illumination conditions. FINCH’s ability to reconstruct spatial images computationally from holograms revolutionized incoherent light imaging; however, it has always been limited by fixed imaging properties once recorded. The novel method, termed Post-Engineering of Axial Resolution in FINCH, or PEAR-FINCH, marks a paradigm shift by enabling post-recording adjustment of focal depth, widening the operational capacity of the technology.

A significant advantage of PEAR-FINCH is its capacity to maintain both high image quality and signal-to-noise ratio during the two-step computational reconstructions. This ensures that the enhanced depth of focus does not come at the cost of image clarity, a common trade-off in many imaging methods that attempt to increase focal depth artificially. Achieving a fivefold increase in depth of focus compared to standard FINCH techniques places PEAR-FINCH as a superior tool for detailed biological imaging, especially in specimens with intricate spatial structures.

One of the technical highlights of this method is its robustness under diffusive illumination — the kind of scattered light typically found in real biological samples. Conventional holography often struggles in such conditions due to loss of contrast and resolution; PEAR-FINCH’s computational sophistication tackles these challenges, making it exceptionally well-suited for real-world biological and biomedical microscopy applications where light scattering and diffusive effects are unavoidable.

The flexibility offered by PEAR-FINCH is unmatched. Researchers now have the unprecedented capability to fine-tune the axial resolution and depth of focus after the hologram recording stage, granting a new realm of adaptability. This flexibility means scientist can tailor imaging parameters according to the requirements of individual samples or experiments without needing to repeat data acquisition—saving time and resources while enhancing scientific precision.

Beyond fundamental research, the implications of this technology extend to medical diagnostics, drug discovery, and other fields that demand intricate 3D imaging under varied and often challenging light conditions. By enabling adaptive and intelligent microscopy, PEAR-FINCH brings researchers closer to the next generation of microscopes that actively respond to and optimize for the imaging challenges presented by complex biological samples.

The research team’s findings were meticulously documented in the Journal of Physics: Photonics, illustrating the profound capabilities and applications of the PEAR-FINCH method. The study not only details the algorithmic framework and optical configuration but also presents rigorous experimental evidence validating the system’s performance across a variety of imaging conditions.

“This technology represents a new standard in holographic imaging,” Gopinath explains. “By facilitating extensive control over imaging properties post-capture, PEAR-FINCH surpasses conventional imaging systems and opens up new investigative possibilities that were previously unattainable.” Such advancements signify a move toward smarter, more precise, and user-driven microscopy platforms.

As microscopy continues to evolve, the intersection of optics with computational methods is proving extremely fruitful. PEAR-FINCH stands as a testament to how these interdisciplinary approaches can overcome physical limitations and enhance image capture for scientific advancement. Future explorations may expand the method’s capabilities further, integrating machine learning and real-time processing to create fully autonomous, self-optimizing imaging systems.

This pioneering work elevates the potential of 3D microscopy, particularly in biological contexts, where observing living organisms or complex tissues in their native state with high fidelity is essential. The ability to manipulate image acquisition and reconstruction post hoc provides researchers with a powerful tool to uncover subtle structural and functional details otherwise masked by traditional techniques.

The University of Tartu’s innovation heralds a transformative step towards more adaptive and intelligent microscopy systems. These developments are set to propel numerous scientific domains forward, providing new insights into biological complexity, improving experimental efficiency, and refining the understanding of intricate three-dimensional structures.

Subject of Research: Not applicable
Article Title: Axial resolution post-processing engineering in Fresnel incoherent correlation holography
News Publication Date: 26-Jan-2026
Web References: https://iopscience.iop.org/article/10.1088/2515-7647/ae38ae
References: University of Tartu Institute of Physics, Journal of Physics: Photonics
Image Credits: Author: Shivasubramanian Gopinath

Keywords

3D holography, computational imaging, PEAR-FINCH, FINCH, depth of focus, holographic microscopy, biological imaging, axial resolution, incoherent light imaging, post-processing imaging, optical imaging advancements, University of Tartu

Tailoring a garment to perfectly match the unique contours and movements of an individual has traditionally required artisanal skill, intensive labor, and custom fittings. Such a process remains inherently unscalable. However, the team’s inventive use of machine embroidery, widely available in industrial and hobbyist contexts but seldom considered for mechanical function, challenges this status quo. By controlling how threads are arranged and connected, they transform embroidery from a purely aesthetic element into a programmable mechanical system capable of encoding complex stretch patterns within fabric.

Taking bioinspiration from human skin, the researchers drew parallels between fibrous biological tissues and woven textiles. Skin’s anisotropic elasticity—that is, its varying stretch properties depending on direction—is governed by the density and orientation of collagen fibers. Wrinkles themselves are a visible testimony to fiber packing, while subtler collagen waves enable additional, nuanced stretch. Mimicking this biological principle, the team conceptualized textiles as fibrous metamaterials whose mechanical behaviors could be tuned via controlled thread architecture.

Central to their approach is the concept of the “fibrous spring”—short zigzag stitches embroidered onto elastic fabric using inelastic polyester thread. Unlike a straight seam that offers no stretch, these zigzag stitches initially exhibit slack and will extend elastically until the thread straightens. By precisely adjusting the amplitude of each zigzag, the researchers dictate the length of slack, and thus the local stretch limit, effectively encoding how much a particular patch of fabric can extend before becoming taut.

This methodology gains further sophistication through the tessellation of fibrous springs into a triangular mesh. Triangles, by their geometric nature, resist distortion without side elongation, making them a perfect base unit for encoding directional mechanics. Embroidery machines are programmed to stitch these triangular spring units in a single pass, with newly embroidered loops knotted firmly around preceding threads. This knot architecture guarantees durability, preventing unravelling and enabling fabric integrity even under repeated mechanical stress.

To bridge artistry and functionality, a custom Python library converts digital images into stitch commands that the embroidery machine can execute. Remarkably, the three color channels of a conventional raster image—red, green, and blue—serve as intuitive inputs to define the mechanical properties of each triangular unit within the fabric mosaic. Designers can thus ‘paint’ zones of varying stretchability using ordinary graphic design software, marrying creativity with engineered performance and significantly simplifying the design-to-production workflow.

The level of control this technique affords is impressive; with a resolution down to 7 millimeters per fibrous spring unit, embroidered textiles can replicate the intricate mechanical behavior of human skin across large surfaces. Thousands of these springs can coalesce into a complex yet coherent fabric that locally adjusts elasticity while collaboratively modulating deformation and load distribution. This grants garment designers unprecedented ability to tailor localized stretch and rigidity, essential for comfort, mobility, and durability.

Directional stretchability is particularly critical in wearables, where tissues like synthetic and natural leather often fail to reproduce the anisotropic compliance of skin. Leather’s tanning process obliterates its native mechanical cues, enforcing homogeneous stiffness rather than dynamic flexibility. In contrast, embroidered fabric engineered by this method acts as a compliant “second skin,” offering a functional mimicry of living tissue’s mechanical anisotropy. This marks a departure from purely cosmetic textile innovation into the realm of functional biomimicry.

A striking proof of concept is embodied in a footwear prototype fabricated from a single embroidered fabric panel containing over a thousand unit cells and nearly twenty thousand stitches. The shoe required minimal post-embroidery assembly, yet conformed naturally to the wearer’s foot, snugly accommodating the heel without sag and restraining unnecessary toe torsion while allowing flexion. Such tailored fit has meaningful implications for reducing injuries in high-coordination sports and physically demanding occupations.

Embedded within this embroidered mosaic is a physical neural network architecture, whereby each knot and node acts as a basic information-processing element. While each fibrous spring operates independently in terms of mechanical stretch, their interconnectedness yields emergent behavior, enabling rapid, responsive adjustments to external stimuli. The prototype shoe demonstrated real-time sensing of foot-ground forces, adapting gait instantaneously based on preprogrammed stitch instructions embedded in the fabric itself.

This integration of hardware and software blurs traditional boundaries, marrying visual aesthetics and textile design with functional programming. The embroidered stitch pattern serves as both an artistic statement and a tangible, legible ‘code’ for the fabric’s mechanical responses—visible to the eye and tactile to touch. According to the researchers, this fusion exemplifies how scientific innovation can manifest as beauty, where the underlying ‘program’ lies hidden in plain sight, encoded seamlessly into the garment rather than imposed as a foreign add-on.

Beyond immediate applications in footwear, this embroidery-driven approach holds promise for mass customization in sportswear, orthopedic supports, and diverse wearable technologies. Its scalability arises from reliance on standard commercial embroidery equipment and accessible textile materials, demanding mere software updates to impart complex mechanical functionalities. The ability to engineer graded, anisotropic stretch at millimeter-scale resolution opens new horizons for personal comfort, adaptive performance, and even interactions between humans and robotic systems.

Moreover, the integration of this approach into wearable robotics offers potential safety and social acceptance benefits by grounding artificial devices in physically intuitive behaviors. By embedding mechanics directly into fabric, robots and AI interfaces can interact with the human body in ways that feel natural and responsive. The innovation thus transcends garment design, gesturing towards a future where intelligent textiles form a foundational interface between biological and machine systems.

In conclusion, the University of Tartu’s researchers have transformed embroidery from a purely decorative craft into a sophisticated tool for multifunctional textile engineering. Their method encodes mechanical properties into fabric at high resolution, enabling personalized, scalable wearables that move and adapt with the body. This advancement could herald a new era in the textile industry, where a simple software update empowers designers and manufacturers to create garments that are not only beautiful but also highly functional and intelligent.

Subject of Research: Not applicable

Article Title: Textile Encoding Inspired by Langer Lines via Elastically Graded Embroidered Tessellations

News Publication Date: 11-Sep-2025

Web References: DOI 10.1002/adma.202500959

Image Credits: Leonid Zinatullin and Indrek Must

Keywords

Embroidery, Textile Mechanics, Wearable Technology, Fibrous Metamaterials, Stretchable Fabrics, Biomimicry, Anisotropic Stretch, Physical Neural Networks, Dynamic Garment Fit, Machine Embroidery, Textile Engineering, Personalized Wearables