A Method Based on Area Histogram for Coarse Carbides Classification in High Speed Tool Steel

Manual methods are used to perform the classification of coarse carbides in high speed tool steel these days, so a quantitative standard is demanded necessarily to achieve the automatic classification. In order to obtain the standard and further perform processing and analysis of carbide metallograph, area histogram used in image retrieval is first introduced to metallographic analysis field. Average area histogram(AAH)and prominent area histogram?PAH? of different levels high-speed steel carbide spectrum are computed, and then the relationship between AAH&PAH and the distribution of carbide is found. Because a large number of small carbides on the calculation of AAH generated interference, the calculation result is not satisfactory so filter coefficients is put forward and used to remove the impact of small pieces of carbides on analysis results. Filter coefficients is further discussed by means of experiments results and AAH&PAH which are regarded as a quantitative standard of the automatic classification are calculated based on the best filter coefficient .