提出一种基于保相位变换(PHOT)的纹理抑制方法对花纹板表面进行处理,抑制规律纹理,突出非规律纹理即缺陷区域.首先对花纹板图像进行预处理,主要包括降采样、亮度补偿及水痕标记,对预处理后的图像进行FFT变换,幅度均一化及FFT反变换,得到仅保留相位信息的频域图像.对得到频域图像求梯度以放大缺陷和正常区域的差异,并采用马氏距离对梯度图进行二值化操作.最后对二值化图像进行包括缺陷重构和去水痕在内的后处理操作,从而实现花纹板表面缺陷检测.经测试,该算法对花纹板表面存在的花纹残缺、花纹错乱、翘皮和油斑四种缺陷都有较好的检出效果.

针对镀锡带钢表面差厚打印线在表面检测系统中的识别和剔除,以机组生产信息进行差厚打印线信息解析,结合图像处理功能有效地将差厚打印线与表面缺陷进行区分和识别.首先,生产系统将含有打印线模式的相关信息通过L3传送给带钢表面检测系统;然后,检测系统解析打印线模式,并综合图像处理结果进行打印线过滤.该技术已在宝钢镀锡带钢表面检测系统中得到成功应用,现场3年多时间的应用表明,该技术可以有效实现差厚打印线在表面检测系统中的检测.

A defect detection and measurement method was proposed for detecting dots and cracks on the surface of strip steel.The method includes the following steps: first,the digital image of the sample is filtered,binarized,thinned and deburred;then dot detection is conducted to the binary image and crack detection to the skeleton image;finally,after computation is made within the defect area,the number and area of dots and the number,length and direction of cracks are obtained.Test results proved the effectiveness of this method in detecting and measuring strip steel defects.

Burrs removing is a classic problem in image processing,and is also the key point of accurate calculation of strip steel crack length.A given length threshold is needed in traditional burrs removing algorithms need,however it is not feasible for crack measurement of strip steel as both long and short cracks may exist in one image.We apply the idea of Local Binary Pattern to identify each skeleton point as nodes,ends or ordinary points.Firstly most burrs are removed according to the relationship of burrs,nodes and ends.Then the shortest branches of remaining nodes would be removed and a skeleton image without burrs is obtained.The experimental result proves that it is an effective burrs removing algorithm which can make the skeleton express the object's geometric structure accurately without fixed length threshold.

FS-Bagging,a new feature selection based ensemble classifier algorithm,is proposed to cope with common classifier problems elicited by large attribute sets.The new classifier algorithm includes an application of both ReliefF algorithm and Pearson algorithm in an initial step,where a sub-optimal attribute set is derived from the original feature set.Subsequently,the sub-optimal attribute set is subjected to a random selection process,in which a number of feature subsets are grouped according to Bagging algorithm.The feature subsets and their corresponding training data sets are further used to construct a group of classifiers,and the final classification is achieved by simple majority voting results of each individual classifier within this ensemble.The feature subsets and their corresponding training data sets are further used to construct a group of classifiers,and the final classification is achieved by simple majority voting results of each individual classifier within this ensemble.A test on this new classifier with 18 main types of surface defect in strip steel production(data retrieved from a surface inspection system in a Chinese steel company)shows positive improvement over classical Bagging classifiers in terms of efficiency and accuracy in classification.