摘要
针对Eitz线画草图识别算法在提取局部特征时没有考虑像素的空间位置信息、分类阶段未选取有效核函数等问题,本文提出一种融合位置信息的类hog局部特征提取算法,旨在更好地区分不同位置上具有相似梯度方向分布的特征;其后采用词包模型对图像进行描述.分类阶段采用直方图交核作为核函数,应用SVM分类器进行分类识别.实验结果表明,本文所提取的局部特征数量少、维数低,有效地降低了运算量.同时还能在大多数情况下提高线画草图的识别率.
In Eitz algorithm, neither spatial location information nor effective selection of kemel functions for classification is consid- ered. To overcome these drawbacks, we propose a hog-like feature extraction algorithm using location information. The proposed al- gorithm aims to make more distinctive the local features with similar gradient distribution but in different positions. A bag-of-features representation is used to define the image. In classification process, we employ SVM classifiers with histogram intersection function as the kernel function. The experiments show the proposed algorithm reduces the computational complexity by extracting fewer local features with lower dimensions, and improves classification accuracy in most cases.
出处
《小型微型计算机系统》
CSCD
北大核心
2014年第8期1897-1900,共4页
Journal of Chinese Computer Systems
基金
福建省自然科学基金项目(2012J01262)资助
福建省高校产学研重大专项(2010H6012)资助