Study on the key technology of spectral reflectance reconstruction based on a single pixel detector. (20th November 2017)

• Record Type: Journal Article
• Authors: Li, Bei; Zhang, Haojun; Zhang, Leihong; Kang, Yi; Zhan, Wenjie; Yi, Wenjuan; Zhang, Dawei
• Is Part Of: Laser physics letters. Volume 14:Number 12(2017:Dec.)
• Publication Date: 2017-11-20
• Subjects: single pixel detector -- spectral reflectance reconstruction -- principal component analysis -- compressed sensing algorithm; Lasers -- Periodicals; Physics -- Periodicals; 621.366
• DOI: 10.1088/1612-202X/aa8cde
• Languages: English
• ISSNs: 1612-2011

Abstract

Abstract: In order to realize an accurate reconstruction of images, this article explores image reconstruction from a spectrum point of view. Combined with a principal component analysis method and compressed sensing algorithm, this article puts forward a spectral reflectance reconstruction algorithm based on a single pixel detector, and compares the different modulation matrixes for DMD in the experiment. From subjective evaluation and three objective evaluation methods: color difference, peak signal-to-noise ratio and structural similarity, we compare and analyze the reconstruction effects of two modulation methods: random modulation and Hadamard modulation. The experimental results show that the color difference of the spectral reflectance reconstruction method of a single pixel detector based on random modulation is smaller, the similarity of the structure is higher, and the peak signal-to-noise ratio is greater than Hadamard modulation. To sum up, the spectral reflectance reconstruction algorithm of a single pixel detector based on random modulation is better than Hadamard modulation, which is more conducive to an accurate reconstruction of spectral reflectance.