Single corn seed moisture content detection method based on near infrared hyperspectrum

Abstract

The invention provides a single corn seed moisture content detection method based on near infrared hyperspectrum, which comprises the following steps: 1) collecting N corn seed samples with different moisture contents in the same batch, and measuring moisture content data yi and hyperspectral image data xi of the samples; 2) preprocessing the hyperspectral image data xi to obtain preprocessed spectral image data x'i; 3) extracting a characteristic spectrum of the preprocessed spectral image data x'i and texture features of the hyperspectral image data xi, and splicing the extracted characteristic spectrum and texture features to obtain a spectral image fusion matrix Pi; 4) dividing the sample set into a training set and a test set, and constructing a moisture prediction model by using a random forest algorithm; and 5) predicting a to-be-detected sample: performing adaptive weighted prediction on the moisture content of the to-be-detected sample according to the random forest model. The near-infrared hyperspectral technology is utilized, spectrum and image data are fused, a single corn seed moisture content detection model is constructed, and the method has important significance on application of the corn precision sowing technology.

Description

technical field [0001] The invention relates to the field of moisture detection of corn seeds, in particular to a method for detecting moisture content of single corn seeds based on near-infrared hyperspectral. Background technique [0002] Corn is currently the crop with the most extensive planting area and the highest yield in my country. It is not only the "king of feed", but also an important industrial raw material. In recent years, with the introduction of "single seed sowing" corn seeds by the American Pioneer Company and the introduction of the mechanized single seed precision sowing technology of corn in developed countries such as Europe and the United States, the traditional corn sowing mode in my country has undergone profound changes. Compared with the traditional sowing mode, the single-seed precision sowing technology is more conducive to mechanized operation, labor-saving, seed-saving, and high-yield, but it puts forward higher detection requirements for each...

AI Technical Summary

Problems solved by technology

In 2010, my country issued the national standard GB / T24900-2010 [12] for the determination of corn moisture content by near-infrared method for the determination of corn seed batch moisture, which shows that near-infrared spectroscopy technology has practical application feasibility in the field of corn seed batch moisture detection, but at present There is no method or standard for near-infrared hyperspectral detection of single seed moisture

Images