摘要
有效去除图像特征中的冗余是图像分类研究领域的一个重要内容。在SPM(Spatial Pyramid Matching)图像分类算法的基础上,结合主成分分析方法(Principal Component Analysis,PCA),提出了一种新的PcSPM算法。该方法能在多种尺度上提取图像金字塔直方图主成分,可减少特征冗余,并将其应用于图像分类。实验表明,该方法能够有效去除图像特征中的冗余,提高了图像分类的精度。
Effective elimination of redundancy in image features is a major content in the research field of image classification.Based on the Spatial Pyramid Matching(SPM)image classification algorithm,and by integrating with principal component analysis(PCA)method,this paper proposed a new PcSPM algorithm.It is able to extract pyramid histogram principal components of image on multiple levels,and reduce feature redundancy and be applied in image classification.Experiment shows that this method is capable of effectively eliminating redundancy in image features and improving the accuracy of image classification.
作者
李青彦
彭进业
LI Qingyan;PENG Jinye(School of Electronics and Information,Northwestern Ploytechnical University,Xi'an 710072;School of Information and Technology,Northwest University,Xi'an 710127)
出处
《微型电脑应用》
2020年第2期17-19,共3页
Microcomputer Applications
关键词
图像分类
SPM
特征降维
主成分分析
词袋算法
Image classification
SPM
Feature dimensionality reduction
Principal component analysis
Bag of Word(BoW)
作者简介
李青彦(1982-),男,菏泽人,博士研究生,研究领域:数字图像仿真处理;彭进业(1964-),男,娄底市人,教授,博士生导师,研究领域:图像处理与模式识别。
