China
Africa
Explore chapters and articles related to this topic
Introduction to Hyperspectral Satellites
Published in Shen-En Qian, Hyperspectral Satellites and System Design, 2020
Both dispersive element based hyperspectral imagers (Section 1.2.1) and spectral filters based hyperspectral imagers (Section 1.2.2) require a scanning to sequentially generate a hyperspectral datacube by either whiskbroom (point-scanning), pushbroom (line-scanning), or wavelength scanning (using spectral filters). A snapshot hyperspectral imager generates a hyperspectral datacube during a single integration time of a detector array. No scanning is required. It incorporates specialized components to distribute a scene's 3D spatial-spectral information onto a 2D detector array. Snapshot hyperspectral imaging typically requires detector arrays with a high number of pixels. The elimination of moving parts means that motion artifacts can be avoided.
Deep learning approach for hyperspectral image demosaicking, spectral correction and high-resolution RGB reconstruction
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2021
Peichao Li, Michael Ebner, Philip Noonan, Conor Horgan, Anisha Bahl, Sébastien Ourselin, Jonathan Shapey, Tom Vercauteren
To achieve intraoperative tissue characterisation with HSI in real-time, snapshot cameras are more suitable as they can capture hyperspectral cube data in real-time (Ebner et al. 2021). A common type of snapshot camera uses a snapshot mosaic system to acquire the entire hyperspectral cube instantly without the need of a scanning mechanism. The refined pixel filter array, arranged similarly to the colour filter array on the RGB sensor, allows the snapshot camera to acquire a maximum of different spectral bands in a single exposure (Geelen et al. 2014). Other snapshot hyperspectral imaging approaches, such as coded aperture snapshot spectral imaging (CASSI) (Wagadarikar et al. 2008) and micro-lens-based acquisition, have been proposed. In general, the downside of snapshot acquisition is that it sacrifices spatial and spectral resolution to achieve fast data acquisition speeds. Figure 1(b) illustrates the relationship between a high-resolution hyperspectral cube and a snapshot mosaic image as a simplified example. An snapshot image is composed of a large number of individual blocks following mosaic patterns. The snapshot on the right of Figure 1(b) is an example of a single block captured by the sensor array.