基于视觉词袋模型的图像分类改进方法

摘要

文中主要基于视觉词袋（BOVW， Bag-Of-Visual-Words）模型对图像进行分类处理，并对传统视觉词袋模型存在的不足进行了改进，提出了一种基于视觉词典的权重直方图来表达图像，采用优化的k-means聚类算法（k-means+）用于视觉词典的构建，代入KNN（K-Nearest-Neighbors）分类器进行分类。通过对Caltech 101和Caltech 256这两个经典数据库进行实验，实验结果表明该改进方案较传统方法提高了分类的正确率。%The Bag of Visual Words (BOVW) model is applied to object classification in this paper, and the shortcomings of the traditional bag of visual words model was improved. This paper mainly presents an optimized k-means (k-means+) cluster-ing algorithm to construct the visual dictionary and a new histogram weighted representation (WR) for images to improve the accuracy of images classification based on K-Nearest-Neighbors (KNN) classifier. The experiments are carried out based both Caltech 101 and Caltech 256 database. The results show that the proposed method performs better than the traditional method and the state of the art.