欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
以多聚焦图像为研究对象，提出一种新的基于多目标优化与熵成分分析的融合方法。源图像经小波分解后分为低频和高频两部分。针对低频信息采用权重系数法融合。将灰色关联分析运用于VEQPSO迭代优化中，用来确定最佳融合系数。源图像分窗体，每个窗体提取特征向量，利用模糊C-均值聚类算法（FCM）在特征空间上分割图像。将窗体进行熵成分变换，构建区域的熵主成分向量，该向量即反映了区域的灰度信息又体现了熵信息。基于熵成分向量设计融合策略更为合理可靠。最后，利用小波逆变换得到融合图像。研究表明，提出的方法具有良好的融合特性。
英文摘要：
Aiming at the multi-focus images, a new method of image fusion based on VEQPSO and entropy component analysis is proposed. Wavelet transform is applied to input images and low frequency parts and high frequency parts are obtained. Average weight fusion method is used for low frequency parts. The weight is defined by VEQPSO based on grey relation analysis and multi-object functions are selected according to fusion purposes. The mean image of source images is partition to windows. The feature vector of every window is extracted. The Fuzzy C-Means clustering algorithm（FCM） is used to segment image in order to obtain the region. Entropy component transform is carried out for the window, and then the entropy component vector of the region is gotten, which reflects the gray information and the entropy of the region. The fusion strategy of high frequency parts is more reasonable and reliable based on entropy component vector. Finally, fused image is obtained by making reverse wavelet transform. The evaluation results indicate that the proposed image fusion method is effective.
同期刊论文项目
同项目期刊论文
An optimal symmetrical null space criterion of Fisher discriminant for feature extraction and recogn
期刊信息
位置：