提出一种用于干涉型光纤陀螺仪(IFOG)装配质量控制的内窥镜系统设计及相关图像分析方法。同时采用1.0mm与2.4mm工业内窥镜系统采集IFOG内部图像与视频数据;并利用多自由度运动控制平台调节内窥镜工作姿态以保证图像的顺利采集。在此基础上,设计了基于db4小波基的图像增强方法,基于灰度共生矩阵的图像纹理特征估计方法以及图像几何形状特征的计算方法等,用以进行内窥镜图像的分析。通过上述系统与方法的应用,能够有效检测出IFOG产品在装配过程中出现的多余物颗粒、光纤断纤或异常胶粘等质量问题,有效提高了IFOG产品的装配质量。 An endoscope system design and related image analysis method for the quality control of interference type optical fiber gyroscope (IFOG) was proposed. At the same time using 1.0mm and 2.4mm industrial endoscope system to collect IFOG internal image and video data; and multi-degree of freedom motion control platform to adjust the endoscope posture to ensure the smooth image acquisition. On this basis, the image enhancement method based on db4 wavelet basis, the image texture feature estimation method based on the gray level co-occurrence matrix and the calculation method of the image geometric shape feature are designed for endoscopic image analysis. Through the application of the above system and method, quality problems such as extraneous particles, fiber broken fiber or abnormal glue appearing in the assembly process of the IFOG product can be effectively detected, and the assembly quality of the IFOG product is effectively improved.