期刊名称：International Journal of Signal Processing, Image Processing and Pattern Recognition
印刷版ISSN：2005-4254
出版年度：2016
卷号：9
期号：6
页码：55-70
DOI：10.14257/ijsip.2016.9.6.06
出版社：SERSC
摘要：Current coins often get damaged in the current environment, such as smudginess, discoloration, corrosion, mutilation and attrition. When the damage comes to a certain degree, the continuous use of coins will be impacted. In this paper, relevant identification technologies to mutilation and corrosion which are two common defects in circulation are investigated. At first, an auto identification and quantization algorithm about mutilation area of coin surface, based on Freeman boundary tracing is designed. On the basis of binarization, the algorithm accomplishes boundary tracing and extraction and area calculation automatically after the seed point is determined which is an arbitrary point in extracting mutilation area. Secondly, a clustering method based on HIS color model is developed for auto identification of corrosion area and calculation of corrosion area and average corrosion degree. At last, the validity and practicability of the algorithms are proved by a large number of experimental results.
