Spectral imaging technology is a new multi-dimensional information acquisition technique that combines imaging and spectroscopy. It can capture a data cube composed of two-dimensional spatial information and one-dimensional spectral information of the target. Through subsequent data processing, spectral curves of different ground objects can be extracted.

Classification of Spectral Imaging Technology

Spectral imaging originated in the 1980s, evolved from multispectral remote sensing imaging. Benefiting from its powerful information acquisition capability, it has developed rapidly with diversified technical branches and continuous iteration of imaging spectrometer products.

There are multiple classification criteria for spectral imaging. According to different spectral splitting principles, it is mainly divided into dispersive type and interference type. Both dispersive and interferential spectral imaging systems acquire 2D spatial and 1D spectral information via push-broom or whisk-broom scanning. They impose strict requirements on platform stability and capture spectral data of all wavebands in a single exposure.

For filter-based spectral imaging solutions — whether adopting multiple filters to capture multi-wavelength images in parallel or switching filters sequentially — the exposure time must be reasonably configured according to the system spectral response, so as to maximize the signal-to-noise ratio.

Dispersive Prism Spectroscopy

The dispersive prism is the most common and simplest spectral splitting component in spectral imaging, and its typical application in spectrometers is illustrated in the figure above. The entrance slit is located on the front focal plane of the collimating system. After being collimated, incident light passes through the prism and is focused by the imaging system, forming wavelength-separated slit images on the focal plane detector.

Diffraction Grating Spectroscopy

The working layout of diffraction gratings is consistent with that of dispersive prisms. The entrance slit is placed on the front focal plane of the collimator. Collimated incident light is diffracted by the grating and projected onto the focal plane detector in order of wavelength.

In another configuration, the grating is placed in a divergent beam. Light from the entrance slit directly irradiates the grating without collimation. Diffraction generates spectral virtual images of the target slit, and the imaging system projects wavelength-resolved signals onto different positions of the area array detector. This design has been adopted in the conceptual design of tactical remote sensors for the OrbView‑4 satellite.

At present, most mature airborne and aerospace-borne dispersive spectrometers worldwide are grating-based, including AVIRIS (Jet Propulsion Laboratory, USA), CASI (Canada), AISA (Finland), and MODIS spectral radiometers.

Binary Optical Element Spectroscopy

Binary optical elements act as both dispersion and imaging components. By scanning the imaging range of selected wavebands along the optical axis with a monochromatic area array detector, each scanning position corresponds to the imaging area of a specific wavelength.

Similar to conventional lenses, binary elements converge incident light based on diffraction principles. The chromatic aberration effective focal length generated by diffraction is inversely proportional to wavelength.

Different from prisms and gratings that disperse light perpendicular to the optical axis, binary optical elements realize axial dispersion. For spectrometers adopting this structure, spectral resolution is determined by detector size, featuring compact structure and high diffraction efficiency.

Acousto-Optic Tunable Filter (AOTF) Spectroscopy

The Acousto-Optic Tunable Filter (AOTF) is a novel dispersive component, consisting of an acousto-optic medium, transducer array, and acoustic termination. Based on the acousto-optic diffraction principle, when polychromatic light incidents on the medium at a specific angle, acousto-optic interaction diffracts incident light satisfying momentum matching conditions into two orthogonal monochromatic beams on both sides of the zero-order light.

Changing the radio frequency (RF) signal frequency adjusts the wavelength of diffracted light. Continuous and rapid tuning of RF frequency enables fast spectral scanning within the operating wavelength range.

Interferential Imaging Spectrometer

For dispersive spectrometers, spectral resolution is inversely proportional to entrance slit width. To improve spectral resolution, the slit must be narrowed, which reduces luminous flux and results in low detection sensitivity.

With rising technical requirements for spatial resolution, spectral resolution, and weak signal detection capability, traditional dispersive systems can no longer meet advanced application demands. Interferential imaging spectrometers feature inherent advantages such as high spectral resolution and high energy utilization. They have gradually become a research hotspot in spectral imaging to satisfy growing application needs.

Main spectral splitting methods for interferential spectrometers include Michelson interference, triangular common-path interference, and birefringence interference. In recent years, new technologies have been developed to generate polarized light via liquid crystal tunable filters for interference. In addition to dual-beam interference, multi-beam interference-based spectral splitting technologies are also widely studied.

Filter-Based Imaging Spectrometer

Filter-based spectrometers integrate optical filters into the optical path as spectral splitting components, acquiring different spectral channels by filter replacement. This technology electrically tunes the central transmission wavelength. The camera exposes once for each wavelength, the system records two-dimensional image data of the corresponding waveband, and then switches to the next transmission wavelength. The cycle repeats until all waveband images are collected to form the final spectral data cube.

Applications of Spectral Imaging Technology

Spectral imaging perfectly combines spectral analysis and image analysis, possessing both spectral and spatial resolution capabilities. It supports qualitative, quantitative and positioning analysis of measured objects. By identifying spectral differences in surface material composition, accurate target recognition and localization can be realized, enabling extensive applications in substance identification, remote sensing, medical diagnosis and other fields.

The development of spectral imaging has gone through three stages: multispectral, hyperspectral and ultraspectral imaging. Imaging spectrometers can acquire multi-band image data with ultra-narrow waveband width, making them ideal for spectral analysis and ground object recognition. Continuous improvement in spectral resolution delivers refined target spectral information, expanding its applications in military, agriculture, medicine, resource exploration, geological survey and other industries.

Military Applications

Capable of distinguishing ground objects through spectral characteristics, imaging spectrometers excel in fine classification, target detection and change detection, serving as a critical battlefield reconnaissance technology. Spectral images can distinguish real targets from camouflaged objects under natural vegetation backgrounds and rapidly detect small tactical targets in desert environments.

Civil Applications

Spectral imaging initially originated from geological and mineral identification research, especially for the detection of special mineral resources such as alteration rocks. Its scope has gradually expanded to vegetation ecology, marine and coastal water color monitoring, water quality detection, ice-snow research, soil analysis and atmospheric science.

High-resolution spectral images integrate map and spectral data. They can accurately extract spectral features of vegetation growth, quantify chlorophyll and suspended matter in water bodies, and detect chemical water pollution. High-precision spectral imaging has become a global research focus, allowing scholars to quantitatively explore material mechanisms at a microscopic scale.