A block compression and reconstruction method for plant hyperspectral

Abstract

The invention discloses a plant hyperspectrum-oriented block compression and reconstruction method. According to the method, an original plant hyperspectrum is subjected to plant region extraction, and then spatial correlation of heights among small block regions is fully utilized in a data expansion form by adopting block compression, so that expanded to-be-reconstructed data is kept consistent with an original single spectrum curve in data arrangement and the length of the to-be-reconstructed data at a time is increased; on one hand, more data is collected once under a same sampling rate, sothat the successful reconstruction probability is greatly increased; and on the other hand, sufficient spectrum information can be obtained by sampling the original plant hyperspectrum with a lower sampling rate, so that the compression and reconstruction of the plant hyperspectrum can be finished with a lower compression ratio.

Description

technical field [0001] The invention belongs to the technical field of plant hyperspectral image processing, and relates to a block compression reconstruction method for plant hyperspectrum. Background technique [0002] The traditional way to obtain plant physiological parameters is to take plant samples and conduct a series of experiments in the laboratory. This process requires a lot of manpower and expensive experimental equipment. Remote sensing technology, as a rapid and macroscopic monitoring technology for surface resources, has the advantages of objective, non-destructive and real-time information acquisition compared with traditional ground surveys. Quantitative inversion brings new opportunities: hyperspectral data contains richer spectral information of vegetation, which greatly improves the accuracy of inversion of vegetation physiological parameters (such as leaf area index and biomass). The most important thing is that it also makes the remote sensing inversi...

AI Technical Summary

Problems solved by technology

However, the actual experimental results show that the reconstruction effect of the single spectral curve compression method is very good when the compression ratio is 0.25 and above, but the reconstruction effect of this compression method is not good when the compression ratio is 0.2 and below, because The data length of a single spectral curve is limited, and the sampling rate is too low, which will lead to too little data collected and cannot restore the original spectrum well

Images