Adaptive clustering lettuce variety identification method

Abstract

The invention discloses an adaptive clustering lettuce variety identification method which comprises the following steps: step 1, collection of lettuce sample spectrums: collecting lettuce samples ofdifferent varieties by using a spectrograph, and obtaining diffuse reflection spectrums of the lettuce samples; step 2, dimension reduction treatment of the spectrum: reducing the dimension of the spectrum of the lettuce sample by adopting a principal component analysis (PCA) method; step 3, carrying out fuzzy C-means clustering on the lettuce spectrum to obtain a fuzzy membership degree and a clustering center; 4, calculating parameters Lambda of the adaptive clustering method; and step 5, identifying the lettuce variety by adopting a self-adaptive clustering method. The method adopts a nondestructive detection technology, does not need sample learning, has the advantages of high detection speed, high identification accuracy and the like, and can be used for accurately identifying lettuceof a plurality of varieties. Particularly, in the aspect of clustering lettuce spectral data containing noise data, the clustering performance is superior to that of fuzzy C-means clustering. The clustering performance is superior to that of the existing fuzzy C-means clustering.

Description

technical field [0001] The invention relates to a lettuce variety identification method, in particular to a lettuce variety identification method using an adaptive clustering method and Fourier spectrum detection technology. Background technique [0002] Lettuce is one of the main vegetables that people often eat. Lettuce is rich in nutrition, it contains a lot of vitamins, β-carotene, dietary fiber and so on. According to the shape of the leaves, lettuce can be divided into head lettuce, wrinkled leaf lettuce and upright lettuce. Different varieties of lettuce have differences in their external shape and internal quality. How to choose lettuce varieties with high yield and high quality is an important research topic for agricultural science and technology workers. Therefore, it is very important to study a simple, fast and non-destructive identification method for lettuce varieties. necessary. [0003] Near-infrared, mid-infrared, far-infrared, ultraviolet-visible absorpt...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem of noise sensitivity of fuzzy C-means clustering, and the influence of fuzzy inter-class scattering matrix on clustering and the identification of lettuce varieties without learning samples, the present invention designs an adaptive clustering and Lettuce Variety Identification Method Based on Spectral Detection Technology

Images