TABLE 3.

The comparison of classification results from different studies.

Study object	Discriminant model	Data resource	Classification accuracy	Source
Bacterial blight resistant rice seed	CNN	Terahertz absorption spectrum	94.95%
Transgenic rice seed	RF	Terahertz reflection spectrum	95%	Hu et al., 2017
Transgenic rice seed	SVM	Terahertz absorption spectrum	92.08%	Lian et al., 2017
Rice false smut seed	RF-ELM	Near-infrared hyperspectral spectrum	91.07%	Wu et al., 2020
DON contaminated wheat	LDA	Near-infrared hyperspectral spectrum	62.7%	Femenias et al., 2020
Maize seeds	RBFNN	Near-infrared hyperspectral spectrum	91.00%	Zhao et al., 2018