Light-field multi-spectral radiation thermometry
Yinsen Luan, Di Mei, Shengxian Shi
Abstract
This Letter proposes light-field multi-spectral radiation thermometry based on an unfocused light-field camera, which can simultaneously record directions and intensities of incident rays. In this method, the direction information of rays is substituted by radiation spectrums via placing an array of filters in front of camera main lens, such that the image sensor can simultaneously acquire spectrums and intensities of rays. By decoupling a raw multi-spectral light-field (MSLF) image and utilizing traditional multi-wavelength pyrometer algorithms, the scalar field of surface temperature distribution can be achieved. To verify the method, measurement errors of different temperature levels on several typical areas of MSLF images are analyzed. In addition, the validation experiment demonstrates that accurate surface temperature measurement can be achieved with a single lens, single monochromatic image sensor, and just one snapshot in the proposed method.