MVI分块标识颜色特征快速检索仿真
|
摘要
为了提高图像检索的效率,提升多媒体用户对图像检索服务的满意度,针对当前多媒体图像检索方法中存在的检索时间较长以及图像分块特征不明显、分块间的特征差异小,导致检索结果匹配度低的问题,提出了一种基于能量函数的多媒体视觉图像分块标识颜色特征快速检索方法。利用几何分解法对图像进行区域划分,并建立图像子区域标记场;结合高斯分布构建标记场的特征能量函数,通过迭代优化获得最优图像分块。计算各分块像素空间的色彩距离均值,作为图像的颜色特征,并引入一维特征向量来表示。利用距离均值计算得到的图像三原色一维特征向量进行检索,采用图像间的颜色分量距离作为相似性度量,通过加权处理完成图像准确检索。仿真证明,上述方法的的分块效果较好,分块特征明显,检索结果匹配度较高,且检索时间较短。
In order to improve the retrieval efficiency of image and enhance the satisfaction degree of multimedia user for image retrieval service,this article proposed a method of fast retrieval for color feature of multimedia visual image block identification based on energy function. Firstly,the geometric decomposition method was used to divide the image regions and established the image sub-region labeled field. Combined with Gaussian distribution,the feature energy function of labeled field was constructed. Then,optimal image block was obtained by the iterative optimization. In addition,the mean value of color distance of each pixel space was calculated,which was used as the color feature of image. Meanwhile,the one-dimensional feature vector was introduced to denote it. Moreover,the mean value of distance was used to obtain the one-dimensional feature vector of three-primary color. Finally,the color component distance between images was taken as the similarity measure. Thus,the accurate image retrieval was finished by weighted processing. Simulation proves that this method has good segmentation effect and obvious block feature. Meanwhile,the method has higher matching degree for retrieval result and shorter retrieval time.
In order to improve the retrieval efficiency of image and enhance the satisfaction degree of multimedia user for image retrieval service,this article proposed a method of fast retrieval for color feature of multimedia visual image block identification based on energy function. Firstly,the geometric decomposition method was used to divide the image regions and established the image sub-region labeled field. Combined with Gaussian distribution,the feature energy function of labeled field was constructed. Then,optimal image block was obtained by the iterative optimization. In addition,the mean value of color distance of each pixel space was calculated,which was used as the color feature of image. Meanwhile,the one-dimensional feature vector was introduced to denote it. Moreover,the mean value of distance was used to obtain the one-dimensional feature vector of three-primary color. Finally,the color component distance between images was taken as the similarity measure. Thus,the accurate image retrieval was finished by weighted processing. Simulation proves that this method has good segmentation effect and obvious block feature. Meanwhile,the method has higher matching degree for retrieval result and shorter retrieval time.
引文
[1]吴治刚.多媒体视觉图像颜色标示优化匹配仿真研究[J].计算机仿真,2017,34(9):235-238.
[2]李士进,等.基于梯度与颜色信息融合的水文资料图像分割[J].数据采集与处理,2016,31(1):94-101.
[3]刘芳辉,郭慧.综合分块颜色体积直方图和LGWBPs的图像检索[J].图学学报,2017,38(3):388-395.
[4]刘芳辉,等.基于基元自相关图和结构元直方图的图像检索[J].电子技术应用,2017,43(8):115-118.
[5]李凯,等.基于分块的多特征融合变尺度目标跟踪算法[J].电视技术,2017,41(1):6-13.
[6]张旭东,等.结合区域协方差分析的图像显著性检测[J].中国图象图形学报,2016,21(5):605-615.
[7]闵永智,等.基于图像色相值突变特征的钢轨区域快速识别方法[J].交通运输工程学报,2016,16(1):46-54.
[8]张玉兵,宋威.视觉特征的分块加权图像检索方法[J].计算机科学与探索,2017,11(3):468-477.
[9]郑永涛,叶仕通,张钊.基于颜色识别的智能搬运机器人的设计[J].电子设计工程,2017,10(25):78-80.
[10]陈璐宇,周春艳.基于分块颜色直方图和GWLBP的图像检索算法[J].液晶与显示,2017,32(9):755-763.
[2]李士进,等.基于梯度与颜色信息融合的水文资料图像分割[J].数据采集与处理,2016,31(1):94-101.
[3]刘芳辉,郭慧.综合分块颜色体积直方图和LGWBPs的图像检索[J].图学学报,2017,38(3):388-395.
[4]刘芳辉,等.基于基元自相关图和结构元直方图的图像检索[J].电子技术应用,2017,43(8):115-118.
[5]李凯,等.基于分块的多特征融合变尺度目标跟踪算法[J].电视技术,2017,41(1):6-13.
[6]张旭东,等.结合区域协方差分析的图像显著性检测[J].中国图象图形学报,2016,21(5):605-615.
[7]闵永智,等.基于图像色相值突变特征的钢轨区域快速识别方法[J].交通运输工程学报,2016,16(1):46-54.
[8]张玉兵,宋威.视觉特征的分块加权图像检索方法[J].计算机科学与探索,2017,11(3):468-477.
[9]郑永涛,叶仕通,张钊.基于颜色识别的智能搬运机器人的设计[J].电子设计工程,2017,10(25):78-80.
[10]陈璐宇,周春艳.基于分块颜色直方图和GWLBP的图像检索算法[J].液晶与显示,2017,32(9):755-763.