A team of multi-institution researchers has presented a continuously tunable wavelength spatial frequency domain imaging (SFDI) system designed for noncontact fruit optical property quantification, published in Engineering. Traditional multispectral SFDI setups face persistent constraints including limited spectral coverage, low optical transmission efficiency, and reliance on mechanical scanning components, which restrict flexible continuous-spectrum characterization of heterogeneous fruit tissues. The modular instrument developed by the group addresses these constraints through an integrated static optical architecture covering a 450–1040 nm operating range without moving scanning assemblies.

The full system consists of four core modular segments: a monochromatic light generation unit built around a 300 W xenon arc lamp paired with a dual-grating monochromator, a digital micromirror device (DMD) projection module, a visible-enhanced near-infrared imaging camera assembly, and a motorized adjustable imaging platform. A custom software framework built on the Sunny.UI interface synchronizes grating rotation, structured pattern projection, and image capture to deliver automated sequential acquisition across variable spatial frequencies and wavelengths, while snap-in optical connectors support future light source and projection module upgrades to extend spectral detection toward the 2500 nm infrared band. Researchers executed systematic characterization and calibration workflows to validate hardware performance, assessing field of view, spatial resolution, wavelength output stability, and image intensity uniformity via standard optical targets and miniature spectrometers. Uniformity correction using a diffuse white reference board reduces intensity variation across projected sinusoidal patterns to mitigate uneven central illumination artifacts inherent to projection optics.

For ground-truth reference, the research team constructed an integrating sphere (IS) measurement system paired with the inverse adding-doubling (IAD) algorithm to acquire benchmark absorption coefficient (μₐ) and reduced scattering coefficient (μₛ′) values. Solid polydimethylsiloxane (PDMS) optical phantoms with calibrated concentrations of India ink and titanium dioxide were used to establish wavelength-specific linear correction functions for the SFDI device, confirming high linear fitting coefficients for both optical parameters across the full spectral range. Liquid phantom pool tests further quantified the system’s depth-resolved imaging capacity, showing detectable light penetration depths of 3–4 mm under coordinated tuning of wavelength and spatial frequency.

Real-fruit validation trials were conducted on peaches, apples, and pears, with SFDI-derived optical property curves aligning closely with IS reference datasets; average measurement error for μₐ stood at approximately 0.002 mm⁻¹, while μₛ′ error reached roughly 0.02 mm⁻¹. Wide-field optical property maps of peach cross-sections revealed distinct spatial heterogeneity in pigment and water absorption signals from fruit exocarp to mesocarp layers. The team deployed partial least squares regression (PLSR) to build peach firmness predictive models using extracted optical properties, with the μₐ-based model delivering stronger prediction performance than frameworks built on scattering coefficients or combined optical metrics. Comparative analysis against prior SFDI hardware confirmed the new platform’s wider continuous tuning range and competitive measurement precision on tissue-mimicking phantoms. The authors outlined existing limitations tied to projection lens infrared coating transmission loss and lengthy multi-wavelength calibration procedures, alongside follow-up directions including spectral extension to the NIR II window, multimodal imaging fusion, and Monte Carlo light transport simulation to deepen understanding of photon behavior within fruit tissue.

The paper “Design, Characterization, and Application of a Continuously Tunable Wavelength Spatial Frequency Domain Imaging System for Measuring the Optical Properties of Fruits,” is authored by Yuan Gao, Zhizhong Sun, Xuan Luo, Dong Hu, Benhui Dai, Yingjie Zheng, Yibin Ying, Lijuan Xie. Full text of the open access paper: https://doi.org/10.1016/j.eng.2026.01.029. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.