It was successful to de-noise the spectrum signal within visual wave band (350 ∼ 550nm) by wavelet transformation and extract some characteristic wavepeaks such as chlorophyll''s 380nm, 414nm, 437nm and the folic acid 366nm. In the range from 550nm to 2500nm, after de-noising, the biggest error was smaller than 1.47%, while at the peaks of waves the biggest error was smaller than 0.11%. The results showed that the wavelet de-noising used in this work slightly or not affect the other non-noise band. The method is very effective in preprocessing spectral data. More consideration of the chlorophyll spectrum characters is helpful to provide a basis for developing chlorophyll meter which has a Strong anti-interference, high reliability and universality.
展开▼
