The viability of Papaver somniferum seeds is the main forensic evidence against crimes related to the illegal possession of drug seed.In order to achieve rapid non-destructive testing of Papaver somniferum seed activity,hyperspectral image technology combined with seed germination rate was used for the first time in this study to establish the prediction model of Papaver somniferum seed activity,and partial least square regression(PLSR),support vector regression(SVR)and extreme learning machine(ELM)were used to establish the prediction model,respectively.The results showed that the prediction model based on ELM had the best performance and Rp was 0.920 4.Compared with the full-band model,the performance of the predic-tion model was improved by using the spectral data of the feature band selected by ELM in the continuous projection algorithm(SPA).The performance of the prediction model established by ELM in the feature band selected by SPA was the best,and the Rp was 0.928 5.The results showed that the method of detecting pop-py seed activity by hyperspectral image was feasible.The method could detect poppy seed activity nondestruc-tively and quickly,and provide technical support for combating the related criminal activities of illegal cultiva-tion of drug plants.
关键词
罂粟/种子活性/高光谱成像/无损检测/预测模型
Key words
Papaver somniferum/seed activity/hyperspectral imaging/non-destructive testing/prediction model
维普
万方数据