基于优化 ELM 的光纤连接器表面自识别降噪技术

光纤连接器的表面检测属于精密仪器检测，因此工厂环境中的大量灰尘会影响连接器表面的复原效果。然而现有的检测技术运行时间长，对于图像细节的保留能力差，并且难以克服实际工作环境中的干扰。因此提出一种优化超限学习机的自识别降噪技术。首先对于干涉数据进行降维处理；其次，采用AdaBoost算法优化超限学习机对噪声点进行定位；最后通过滤波算法对噪声点位置进行修复。实验得出，基于AdaBoost-Elm的自识别降噪算法具有较高的噪声识别能力，其平均噪声识别率达97.33%。此外，采用基于AdaBoost-Elm降噪算法得到BBS的平均值为131.14,NRIQAVR的平均值为2.61,降噪效果均优于全局滤波算法。最后，通过模拟工厂环境，采用基于AdaBoost-Elm的中值滤波算法在不同光强条件下对重度污染的光纤探头进行3D复原测试，其BBS达到130左右，NRIQAVR低于2.57,对比基于Elm的中值滤波算法具有明显优势。

Optical fiber connector surface self-identification noise reductiontechnology based on optimized ELM

The surface detection of optical fiber connector belongs to precision instrument detection, accordingly, making it possible forthe large amounts of dust in the factory environment that exerts detrimental influence on the recovery of optical fiber connector.Nonetheless, the current detection technology possesses long running time, poor retention ability for image details, and is problematic toovercome interference in the actual working environment. To this end, we propose a self-identification noise reduction technology basedon optimised extreme learning machine. Firstly, the interference data is processed by dimensionality reduction. Secondly, select thedimensionality reduction data as the training data, and use the extreme learning machine optimised by AdaBoost algorithm to locate thenoise. Ultimately, the positions of noise points are repaired by filtering algorithms. The experimental results demonstrate that the selfrecognition noise reduction algorithm based on AdaBoost-Elm is equipped with high noise recognition ability and its ANRR reaches97. 33%. Additionally, the average value of BBS and NRIQAVR based on AdaBoost-Elm noise reduction algorithm are 131. 14 and 2. 61respectively, which are better than global filtering algorithm. In the end, we simulate the factory environment and use mean filteringbased on AdaBoost-Elm to perform 3D restoration test on the sharply polluted fiber optic probe under different light intensity conditions.It is found that its BBS reaches around 130 and its NRIQAVR is lower than 2. 57, which has apparent merits compared with the medianfiltering based on Elm.