Machine recognition method for insect piercing-sucking potential waveforms

The invention discloses a machine recognition method for insect piercing-sucking potential waveforms. The method comprises the steps that firstly, denoising preprocessing is conducted on the collectedEPG waveforms; fractal dimension features and HHT features of the EPG waveform are extracted respectively, and series fusion is performed to form feature vectors; and finally, classifying and identifying the six common EPG waveforms by using a decision tree. The method can effectively solve the problem that the EPG waveform depends on artificial recognition all the time, shortens the analysis time of the EPG signal, and promotes the efficient utilization and intelligent development of the EPG. 本发明公开了一种昆虫刺吸电位波形的机器识别方法，所述方法包括首先对采集的EPG波形进行去噪预处理；然后分别提取EPG波形的分形维数特征和HHT特征，串联融合构成特征向量；最后用决策树对六种常用的EPG波形进行分类识别。本发明能有效地解决EPG波形一直靠人工识别的问题，缩短EPG信号的分析时间，促进EPG的高效利用及智能化发展。