基于小波包和神经网络的血细胞识别方法的研究

Blood Cell Recognition Based on Wavelet Packet Analysis and the Neural Network

根据血细胞信号的特点,提出了一种基于小波包分析和神经网络的血细胞识别方法。该方法首先对血细胞信号进行小波包分解,然后对分解系数进行重构,求得重构信号的能量;然后选取三个能量特征并结合7个时域特征参数构造成特征向量,作为神经网络的输入;最后建立神经网络模型进行训练。实验分析了不同条件下的信号识别情况,结果表明该方法识别效果较好。

In this paper, we present a blood cell recognition method based on wavelet packet analysis and the neural network. The blood cell signals were decomposed, then the discrete wavelet coefficients were reconstructed and energy values were calculated. The energy values together with seven features in time domain were used as the inputs of the BP network. Finally the network was established and trained. The accuracies of recognition on different conditions are discussed too. The experimental results show that the proposed method has a high accuracy of recognition.