Research on Identification Method of Sorghum Varieties Based on Machine Learning

Sorghum varieties are closely related to sorghum yield. To achieve rapid identification of sorghum seed varieties, a sorghum variety recognition model based on machine learning was proposed. The morphological features of three kinds of sorghum seed hyperspectral images are extracted by pruning and segmentation, then data features are analyzed and dimensioned by principal component analysis, finally, SVM classifier is used to identify sorghum seed varieties, and high accuracy is achieved.