基于亚像素区域加权能量特征的多尺度图像融合算法

对矩形和圆形区域中各像素进行亚像素划分,确定各亚像素的权值,得到基于哑像素的综合加权区域能量.融合箅法首先对源图像进行金字塔分解,然后对金字塔的高频细节分量使用基于哑像素加权区域能量特征的融合规则取大,对低频粗糙分量取平均.得到融合图像的塔形分解,最后重构融合图像.仿真结果表明,新算法融合效果较常规的区域能量特征作为融合规则的多分辨率图像融合算法效果更好,从清晰度和熵的评价来看,提高了融合图像的品质.

Multiscale Image Fusion Algorithm Based on Subpixel Weighted Region Energy

Dividing pixels in rectangular and circular regions into subpixels and determining weight of each subpixel the composite weighted region energy of each pixel is obtained in the region. The source images are decomposed into multiresolution representation firstly. Then we select average of the coarse parts and maximal parts of the detail components based on the new weighted region energy as the fusion coefficients. Experimental results show that the proposed method has better fusion effects than traditional image fusion schemes based on the traditional region energy. From the evaluation of image definition and entropy, this method improves the quality of the fusion image.