摘要
在软硬件技术不断更迭的过程中,数字图像作为清晰直观的视觉信息载体,其共享与获取也变得更加便捷,在网络数据中的占有量也越来越大。为了提高图像检索效率,以快速精准的从众多的图像信息中找到感兴趣的部分并进行进一步的处理,提出了一种基于多特征融合的织物图像检索算法。阶段一该算法首先对图像库织物图像进行预处理,再对织物图像进行特征提取并生成对应的特征向量,建立特征数据库,将各向量分别存在各自特征数据库中。阶段二将待检索织物图像进行图像预处理,再提取图像各特征并生成对应的特征向量,提取数据库中各特征信息,计算特征向量间的距离,并根据距离对图像库中图片打分,最后根据特征权重进行多特征的融合返回最相似的图片。实验结果表明,该算法可以更为准确的检索出图像库中的织物图像。
In the process of software and hardware technology constantly changing,digital image as a clear and intuitive visual information carrier,its sharing and acquisition has become more convenient,and its share in network data is also increasing.In order to improve the efficiency of image retrieval,a fabric image retrieval algorithm based on multi-feature fusion is proposed in this paper,which can quickly and accurately find the interested part from many image information and further process it.In the first stage,the algorithm first preprocesses the fabric image in the image database,then extracts the features of the fabric image and generates the corresponding feature vectors,establishes the feature database,and stores each vector in its own feature database.In the second stage,the fabric image to be retrieved is preprocessed,and then the image features are extracted and the corresponding feature vectors are generated.The feature information in the database is extracted,and the Euclidean distance between the feature vectors is calculated.According to the distance,the images in the image database are scored,and the most similar images are returned by multi feature fusion according to the feature weight.The experimental results show that the algorithm can retrieve the fabric image more accurately.
作者
李锋
冯益青
LI Feng;FENG Yi-qing(Software Engineering,Donghua University,Shanghai 201620,China)
出处
《新一代信息技术》
2021年第14期41-47,共7页
New Generation of Information Technology
关键词
特征提取
特征融合
织物图像检索
feature extraction
feature fusion
fabric image retrieval
