Position-independent human body action recognition method based on CSI (Channel State Information) and dual-thread convolutional network

The invention discloses a position-independent human body action recognition method based on CSI (Channel State Information) and a double-thread convolutional network. The position-independent human body action recognition method comprises the following steps: 1, collecting CSI action sample data; 2, preprocessing the CSI action sample data; 3, constructing a dual-thread convolutional network; 4, inputting the preprocessed training sample into a double-thread convolutional network for training to obtain a classification model; and 5, inputting the preprocessed test sample into the classification model for human body action recognition. According to the method, action recognition at any indoor position can be realized without providing a sample at a new position or training a model again by a user, and the practicability is relatively high. 本发明公开了一种基于CSI与双线程卷积网络的位置无关的人体动作识别方法，其步骤包括：1、采集CSI动作样本数据；2、对CSI动作样本数据进行预处理；3、构建双线程卷积网络；4、将预处理后的训练样本输入到双线程卷积网络进行训练得到分类模型；5、将预处理后的测试样本输入分类模型中进行人体动作识别。本发明无需用户提供新位置上的样本或再次训练模型即可