Rapid nondestructive high-accuracy method with for identifying abundance degree of nitrogen element in plant leaf

Abstract

The invention discloses a high-spectrum-based rapid nondestructive high-accuracy method for identifying the abundance degree of a nitrogen element in a plant leaf. The method comprises the following steps: performing high-spectral collection on a leaf sample; performing noise-reduced preprocessing on high-spectral data and then performing characteristic extraction on the spectral data; establishing a correction model of the spectrum characteristic and the category of the level standard of the nitrogen element in the plant leaf; collecting high-spectral data of an unknown sample; performing the noise-reduced preprocessing on the spectral data of the unknown sample and then converting a spectrum of the unknown sample to form a characteristic space; substituting the correction model into the characteristic space so as to perform identifying on the level category of the nitrogen element, so as to obtain whether the plant leaf is in nitrogen deficiency or contains the nitrogen element with the normal / excessive amount. According to the technical scheme, a portable high-spectral instrument is utilized to collect the spectrum of the plant leaf; an Adaboost+SLPP characteristic extraction method is utilized to perform the characteristic extraction on the spectral data of the plant leaf; and a KNN (K-Nearest Neighbor) serves as a categorizer, so that the identification predicting accuracy is improved effectively.

Description

technical field [0001] The invention relates to the technical field of non-destructive detection of plant leaf nitrogen level abundance and deficiency indicators, in particular to a method for non-destructive detection of plant leaf nitrogen index level categories based on hyperspectral. Background technique [0002] The non-destructive identification of nitrogen index deficiency in plant leaves is mainly to intelligently judge the current nitrogen deficiency, excess nitrogen, and normal nitrogen levels of plants. It can monitor the growth of plants in real time. Through the detection results, it can not only adjust the supply of nitrogen fertilizers, but also realize Fertilizer management saves resources, reduces soil pollution, and can scientifically guide cultivation work in order to promote healthy and good growth of crops and improve crop yield and quality. [0003] At present, the non-destructive detection methods of nitrogen in plant leaves include spectroscopic metho...

AI Technical Summary

Problems solved by technology

[0005] In view of this, the main purpose of the present invention is to provide a non-destructive detection method for plant leaf nitrogen indicators, which can solve the problems of low model usability, poor anti-interference ability, long detection time, and low detection accuracy in existing spectral non-destructive detection methods. question

Images