欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
本文提出了一套基于灰度直方图和支持向量机的磁环自动分类系统。为了用低维的灰度信息来描述磁环的特征，提出了一套图像处理的算法。将图像从背景分离之后，进行灰度直方图处理来提取灰度特征。接着采用主分量分析法，将灰度统计信息由256维向量降低到20维向量，以这20维向量作为输入，用支持向量机进行分类。最后，经过训练得到最优分类函数，分类正确率达到97．3％。
英文摘要：
An automated classification system for ferrite cores was developed using gray histogram and support vector machine . To characterize ferrite core with low dimensional gray level information, a set of image processing algorithms was developed. After image segmentation from the background, the quantified gray features of the segmented image were extracted using gray histogram. Subsequently, principal component analysis was applied to reduce the gray statistic information from 256-dimensional vectors to 20-dimensional vectors. With the 20-dimensional vectors as the input, SVM classifiers were used for the classification of the ferrite cores. Finally, the optimized classifier was found and it resulted in the best classification accuracy of 97.3 % in the test experiments.
同期刊论文项目
同项目期刊论文
期刊信息
位置：