Simultaneous correction of spectral-line bending across multiple slits in a slit recombination hyperspectral imager with multiple viewing angles

Abstract Simultaneous spectral-line bending correction is required for the central and off-axis slit channels of a multi-angle slit-recombination hyperspectral imager. This study develops a modeling and optimization method using an oppositely oriented dual-material prism group. The model describes spectral-line bending as a function of wavelength, slit channel, and position along the slit. A prism-group merit function guides material screening, and a system-level merit function incorporating distortion guides genetic-algorithm optimization of the prism parameters. Complete-system optimization and numerical evaluation are performed in Zemax for a three-slit design operating over 410–900 nm, using the H-QF3/D-ZLaF81-25 prism group as the sole dispersive element. For the evaluated wavelengths and field positions spanning 410–900 nm, the maximum optical spectral-resolution bandwidths, quantified by the full width at half maximum (FWHM) of the wavelength-domain line-spread function (LSF), are 9.35, 9.43, and 9.75 nm for the + 15°, 0°, and − 15° viewing channels, respectively, all below the 10-nm requirement. The ray-tracing results show that the three slit spectra are geometrically separated on the detector plane. The spectral-line bending and keystone distortion of all slit channels are less than 0.5 effective pixel in the spectral and spatial directions, respectively. At 650 and 900 nm, the sagittal and tangential optical modulation transfer function (MTF) values exceed 0.7 at 25 lp/mm for the plotted field points in all three slit channels. Relative to a reference with two upright prisms, the optimized design reduces the maximum absolute spectral-line bending to 0.0012–0.0024 mm, with an average reduction of approximately 94%. This method provides a design reference for multi-angle hyperspectral imaging systems with low spectral-line bending and geometrically separated slit spectra.