The invention discloses a micro-nano plastic
analysis method based on nano hyperspectral
fingerprint characteristics, and belongs to the technical field of environmental
pollutant detection. The method comprises the following steps: establishing an initial spectrum
library [SL1.0] containing a plurality of plastic polymers; optimizing the spectrum
library through
principal component analysis, extracting characteristic peaks, determining 600-840nm as a plastic
fingerprint region, and constructing an optimized spectrum
library [SL2.0]; preparing an atmospheric
sedimentation sample or a
biological tissue paraffin section; and collecting a hyperspectral image of the sample, and matching the hyperspectral image with the optimized spectrum library in the
fingerprint region by using a spectrum angle matching
algorithm to realize component identification and
spatial positioning of the micro-nano plastic. According to the method, spectral redundancy and
noise interference are effectively eliminated through PCA optimization, the multi-component identification accuracy is remarkably improved by focusing the fingerprint area, the form, component and distribution information of the micro-nano plastic can be obtained at the same time, and a reliable technical means is provided for
environmental monitoring and toxicological research.