基于多特征光谱融合的蔬菜农药残留检测模型

Vegetable Pesticide Residue Detection Model Based on Multi Feature Spectrum Fusion

针对蔬菜农药残留检测问题,提出一种基于多特征光谱融合准则的蔬菜农药残留检测模型.该模型对蔬菜的光谱图像特征进行主成分降维分析,分别基于全光谱和特征光谱进行不同线性判别特性的检测性能分析.最后,针对单一特征检测和多特征融合检测的检测效果进行了充分的仿真分析.实验结果表明,基于多特征光谱融合的蔬菜农药残留检测方法明显提升了农药的检测正确率.

Aiming at the problem of excessive pesticide residues in vegetables,a new model of vegetable pesticide residues detection based on multi feature spectrum fusion criterion is proposed.The model is used to analyze the main components of the spectral image characteristics of vegetables,and the detection performance of different linear discriminant characteristics based on the whole spectrum and the characteristic spectrum is analyzed respectively.Finally,the detection results of single feature detection and multi feature fusion detection are simulated and analyzed.The experimental results show that the multi feature fusion detection method based on this method significantly improves the detection accuracy of pesticides.