摘要
文章对基于多特征提取和改进的SVM分类器的新型图像分类算法进行了理论分析和数值分析。计算机视觉两个基础研究问题是图像对象分类及检测,图像分类方法的核心是利用海量图像冗余数据进行图像特征互补提取,造成图像分类精度较差。本文结合图像分类的特点和综合方法提出了改进的支持向量机(SVM)。该方法可以对图像内容特征进行综合描述,利用主成分分析提取变换特征,消除冗余信息,通过实验结果证明了该算法的有效性和可行性。
In this paper,theoretical analysis and numerical analysis of a new image classification algorithm based on multi-feature extraction and improved SVM classifier are carried out.The two basic research problems of com⁃puter vision are image object classification and detection.The core of the image classification method is to use mas⁃sive image redundant data for complementary extraction of image features,resulting in poor image classification ac⁃curacy.This paper proposes an improved support vector machine(SVM)based on the characteristics of image classi⁃fication and comprehensive methods.This method can comprehensively describe the image content features,use principal component analysis to extract transform features,eliminate redundant information,and prove the effective⁃ness and feasibility of the algorithm through experimental results.
作者
魏彬
Wei Bin(Shaanxi Railway Institute,Weinan 714000,China)
出处
《粘接》
CAS
2021年第3期92-95,共4页
Adhesion
基金
渭南市2019年度重点研发计划项目(2019ZDYF-JCYJ-146)。
关键词
图像识别
深度学习
支持向量机
特征提取
image recognition
deep learning
support vector machine
feature extraction
