图像识别技术在基桩质量检测中的应用研究

在铁路公路桥梁基桩质量检测中，声波透射法被广泛应用。用于评价桩身质量的声学参量主要为声速、声幅、PSD值。波形特征是反映基桩完整性的主要依据。但现行检测分析方法中未考虑波形特征。本文在传统的检测分析基础上采用图像识别技术，将波形特征纳人分析范畴。通过实验室制作了含不同缺陷程度的混凝土立方体试块模拟现场基桩混凝土缺陷，进行声波透射法测试，建立了缺陷程度与波形特征值、声速、声幅的定量关系。工程实例表明，波形特征值综合了声速、声幅以及波在混凝土传播过程中的反射、散射等畸变特征，能客观地反映桩身混凝土质量，与波速、波幅联合使用，提高了分析判定的可靠性。

Study on Application of Image Recognition Technology to Pile Quality Testing

In the testing of bridge pile quality of railway and highway, the sonic transmis- sion method has been widely applied to acoustic parameters, evaluating quality of the pile, and mainly to the acoustic velocity, amplitude and PSD value. Waveform characteristics were the main basis for pile integrity. But waveform characteristics were not considered in current testing and analysis method. In this paper, the waveform characteristics were ana- lyzed using image recognition technology based on traditional testing and analysis. In the laboratory we made the concrete cube test block with different degree of defects simulating pile concrete defects, testing them by acoustic transmission method and established the quantitative relations of degree of defect and waveform characteristic value, the acoustic ve- locity and amplitude. The test shows that the waveform characteristic value integrates the acoustic velocity, amplitude and the distortion characteristics of the reflection and scatter- ing in the transmission of wave, which can objectively reflect the quality of pile concrete with the use of the wave velocity and amplitude.