汇报题目：参加The First World Congress on Condition Monitoring (WCCM 2017) 参会报告

汇报时间：2017年6月27日(星期二) 19:30

汇报地点：兴庆校区西二楼东121会议室

汇报人：徐斌

会议名称：The First World Congress on Condition Monitoring (WCCM 2017)

会议时间：13-16 June 2017

会议地点：The Earls Court ILEC Conference Centre, London, UK.

会议简介：The International Society for Condition Monitoring (ISCM) and the British Institute of Non-Destructive Testing (BINDT) are delighted to invite you to this global event. This congress is of major significance, being the first world event in its field. Ownership and overall control of the event rests with the ISCM and this premier event is being organised in the United Kingdom by BINDT in collaboration with almost all condition monitoring and NDT societies worldwide. This combination of efforts will create one of the largest events of its kind at a truly international level. The event will provide you with a unique opportunity to network with leading academics and industrialists from all over the world in the field of condition monitoring and related areas.

会议交流工作

Oral presentation: An Adaptive Thresholding Segmentation Method of Ferrographic Image for Oil Condition Monitoring

报告人：徐斌

参加论文信息

Title: An Adaptive Thresholding Segmentation Method of Ferrographic Image for Oil Condition Monitoring

Author: Bin Xu, Guangrui Wen, Zhifen Zhang*, Feng Chen

Abstract: Ferrographic image segmentation is of great significance in wear particles recognition and oil condition monitoring. In this paper, an adaptive thresholding segmentation (ATS) methodology was proposed to process ferrographic images for oil condition monitoring. Firstly, the preliminary conceptions of newton interpolation polynomial and difference quotient were briefly introduced, and the grayscale ferrogrphic image was converted into three-dimensional gradation histogram. Then, the slicing operation for three-dimensional gradation histogram was conducted, and the pixels number of each slice was used to plot the curve of slice grayscale-frequency. The two acceptable functions and two acceptable errors were generated from the curve of slice grayscale-frequency. Subsequently, the acceptable threshold can be selected adaptively based on the acceptable functions and errors. Segmentation performance of the ATS method was assessed and validated using a variety of ferrographic images. By comparing with three classical algorithms i.e. Riddle’s iterative thresholding method, Kapur’s maximum entropy method and Otsu method, the results showed that the proposed method possess better performance on both segmentation accuracy and noise resistance. The promising results suggest that the ATS algorithm can be considered as an effective alternative for ferrographic image segmentation.

欢迎感兴趣的同学前来交流！