Spectral Image Formation Process
In this chapter, we describe the image formation process resulting from the transfer of light energy with different wavelengths from the source to the spectral imaging sensor. To commence the chapter, we provide preliminary concepts in radiometry and phot
- PDF / 553,458 Bytes
- 19 Pages / 439.37 x 666.142 pts Page_size
- 96 Downloads / 182 Views
Spectral Image Formation Process
Before proceeding further, it is necessary to provide some background on the image formation process in relation to narrowband spectral images. The concepts involved in the image formation process described in this chapter are fundamental to the formalism in the subsequent chapters. Furthermore, it is important to provide a link between spectral and trichromatic imagery. This relationship further clarifies how the physical aspects of the scene, which can be captured and recovered from spectral imagery, affect the colour perceived by a human observer or that output by a trichromatic sensor. A number of concepts involved in modelling the scene reflectance, such as light scattering and Fresnel reflection theory, are of particular relevance to the physics of polarisation and illumination invariants. The material in this chapter also clearly points out the contributions of separate elements of the physical world, such as geometry, illuminants, and materials, to the subjective colour sensation of a human observer and the colour response of a trichromatic sensor. Moreover, the ability to produce a sensor-dependent colour image from a corresponding spectral image is a highly utilised tool in the subsequent chapters for the purposes of data generation and image display. Therefore, the focus of this chapter is to establish an understanding of the formation process of spectral imagery and its relationship to colour imaging. To commence, we introduce relevant concepts and formulations from the areas of radiometry, photogrammetry, colourimetry, spectral imaging and reflectance modelling. Subsequently, we show how camera simulation and comparison can be effected with respect to the CIE-1931 colour standard. Here, we use the spectral response of the camera to recover RGB colours corresponding to known illuminant and material reflectance spectra. We have compared these RGB values to those computed making use of the colour matching functions proposed by Stiles and Burch (1955, 1959). Recall that, in colourimetry, the aim is to capture and reproduce colours to achieve perceptual accordance between the scene image and the observation by the viewer. The simulation and evaluation of this information is important to the understanding of the relation between the scene and the camera image. Moreover, the accurate capture and reproduction of colours as acquired by digital camera sensors is an active area of research which has applications in colour correcA. Robles-Kelly, C.P. Huynh, Imaging Spectroscopy for Scene Analysis, Advances in Computer Vision and Pattern Recognition, DOI 10.1007/978-1-4471-4652-0_3, © Springer-Verlag London 2013
17
18
3 Spectral Image Formation Process
tion (Brainard 1995; Finlayson and Drew 1996; Horn 1984; Wandell 1987), camera simulation (Longere and Brainard 2001) and sensor design (Ejaz et al. 2006). However, the capture and reproduction of colours that are perceptually consistent with human observations is not a straightforward task. Digital cameras comprise three kinds
Data Loading...
