Polarisation of Light
In this chapter, we provide an introduction to spectro-polarimetric imaging. We show that the radiance transmitted through a linear polariser varies as a sinusoidal function of the polariser angle, i.e. the transmitted radiance sinusoid curves. We review
- PDF / 711,005 Bytes
- 31 Pages / 439.37 x 666.142 pts Page_size
- 5 Downloads / 181 Views
Polarisation of Light
The polarisation of light is a concept describing the distribution of its electromagnetic field at different oscillation orientations in the plane perpendicular to the propagation direction. It has long been a well-studied subject in astronomy (Hall 1951), optics (Born and Wolf 1999; Mandel and Wolf 1995) and crystallography. However, polarimetric imaging is a somewhat recent development, with the early work done by Wolff (1989). Although the human vision system is oblivious to polarisation, a number of animals, such as mantis shrimps, naturally possess a polarisation vision system (Marshall et al. 1991). Biology researchers have also noticed evidence for biophysical mechanisms of polarisation coding in fish (Hawryshyn 2000). With recent advances in camera technologies, these polarisation effects can be captured by devices such as polarimeters and more recently, polarisation cameras (Wolff 1997; Wolff and Andreou 1995; Wolff et al. 1997). In (Wolff 1997; Wolff and Andreou 1995; Wolff et al. 1997), Wolff et al. developed a liquid crystal polarisation video camera with twisted nematic liquid crystals that were electro-optically controlled to replace the need for a mechanical rotation of linear polarisers. The development of these portable, low-cost and fast polarisation camera sensors potentially extends the applications of polarisation imaging to areas such as target detection and segmentation (Goudail et al. 2004; Sadjadi and Chun 2004) and material property recovery (Wolff and Boult 1989). This chapter explores computational models of polarimetric imaging and its applications to the recovery of shape and material properties. We commence the chapter by providing the physical background on polarisation of light as an electromagnetic wave in Sect. 10.1. In particular, we elaborate on the polarisation effects caused by surface reflection, including both diffuse and specular reflection. We describe the related physical processes in detail in Sects. 10.2.1 and 10.2.2. We also draw on the Fresnel reflection theory to formulate a computational model of polarimetric imaging in Sect. 10.3.1. Subsequently, we review techniques concerning the use of polarimetric imaging for the problem of shape and material refractive index recovery in Chap. 11. A. 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_10, © Springer-Verlag London 2013
209
210
10
Polarisation of Light
Notation
Description
Iϑ (u, λ)
Transmitted irradiance at a pixel u and a wavelength λ corresponding to a polariser angle ϑ. Maximum and minimum transmitted values on the transmitted radiance sinusoid (TRS). Unpolarised irradiance. Parallel and perpendicular components of the incident irradiance. Parallel and perpendicular components of the reflected radiance. Parallel and perpendicular components of the transmitted radiance. Electric field vector with components Ex and Ey , wave number k and phase difference ε. Unpolarised and complete
Data Loading...
