Pipeline scene recognition method based on convolutional neural network

The invention provides a pipeline scene recognition method based on a convolutional neural network, and the method comprises the following steps: collecting audio data of a plurality of pipelines in different scenes, and forming a label set of scene types; multiple pipelines under different labels are selected to record audio data, and an original audio file set D is formed; constructing a recognition model; training and evaluating the identification model; and collecting audio data of an unidentified pipeline, performing feature extraction, and identifying the unidentified pipeline by using the model. According to the pipeline scene recognition method provided by the invention, the environment where the pipeline is located can be recognized only by collecting the pipeline audio data through the audio equipment, and further manual measurement is not needed. 本发明提供了一种基于卷积神经网络的管道场景识别方法，包括以下步骤：采集不同场景的多处管道的音频数据，形成场景类别的标签集；选取不同标签下的多处管道录制音频数据，形成原始音频文件集D；构建识别模型；训练并评估识别模型；采集未识别的管道的音频数据并进行特征提取，利用模型对未识别的管道进行识别。通过本申请提供的管