Fig. 1. Schematics of the optical setup and the timing chart. (a) Schematics of the setup, particle is monitored by laser diode 1 and a PD, the presence of particle represents as an electrical pulse, which triggers the laser diode 2 with high power, the backward polarization components of the scattered light are registered by two APD detectors; (b) timing chart

Fig. 2. Measurement results of three materials. (a) The relationship between total backscatter intensity and polarization ratio; (b) the relationship between total backscatter intensity and particle diameter, circles indicate cases where the polarization ratios of the two materials overlap;(c)‒(e)total backscatter intensity to polarization ratio of PEEK, TPU and PMMA

Fig. 3. Comparison of experimental results of binary classification SVM. (a) Raw data of 50 measurements of TPU and PMMA particle polarization ratio; (b) the classification effect of mapping data to other feature spaces (i.e., normalized polarization intensity in the orthogonal direction) and applying linear kernels to classify them

Fig. 4. Comparison of SVM results of ternary classification. (a) The actual classification of multiclass particle types based on ovr; (b) classification of particle types by analyzing the polarization ratio of the particles

Fig. 5. Comparison of the parallel and vertical polarization component acquisition effects of scattered light by different polarization modules. (a) Placing the thin film polarizer directly orthogonally on either side of the intersection surface of the particle stream and the laser beam; (b) directly orthogonal placement of the line grid polarizer on either side of the intersection surface of the particle stream and the laser beam; (c) a combination of an unpolarized beam splitter (1∶1 splitting ratio) and a thin-film polarizer; (d) a combination of an unpolarized beam splitter (1∶1 splitting ratio) and a grid-polarizer; (e) configuration based on a polarization beam splitter and two photodetectors; (f) the sampling intensity and the calculated polarization ratio according to the different configurations

Fig. 6. Experimental results and simulation. (a) Experimental results of polarization ratios when particle size varying from 1‒14 μm; (b) simulated results of polarization ratios when particle size varying from 1‒15 μm

Fig. 7. Experimental results of particle flight time and speed of motion. (a) Scattering pulse of a carbon particle, extraction of the pulse half-maximum full width for estimation of the time-of-flight measurement; (b) time-of-flight distribution of carbon black particles with different particle sizes; (c) movement time distribution of carbon black particles with different particle sizes