金属断口图像处理研究进展

Review of Research on Metal Fracture Image Processing

论述了金属断口图像处理在国内外的发展与研究现状,以及金属断口图像的预处理过程、对图像的特征提取以及模式识别3个部分,并基于空间域及变换域对三部分的方法进行统计分析以及归纳总结。同时,针对算法的运算时间及速率、识别率以及鲁棒性等性能指标,对金属断口图像的识别综合性能进行判断分析和探讨。指出现阶段金属断口图像处理算法存在的不足之处及其进一步的发展方向,这将为断口图像智能诊断技术的发展提供有力的参考。

This article mainly discusses the development and research status of metal fracture image processing at home and abroad.This topic covers three parts of metal fracture image preprocessing,image feature extraction and pattern recognition.Based on spatial domain and transform domain,the three methods are statistically analyzed and summarized.At the same time,according to the algorithm ＇s computing time,speed,recognition rate,robustness and other performance indicators,the comprehensive performance of metal fracture image recognition is analyzed and discussed.At the same time,the shortcomings of metal fracture image processing algorithm and their further development direction are pointed out,which will provide a powerful reference for the development of intelligent image fracture imaging technology.