THT机制耦合高斯模糊的图像融合方案

发布时间：2019-02-15 05:42

【摘要】：为了提高当前红外(IR)与可见光(VI)图像的融合质量,保持良好的可视化细节和突出IR目标,有效减少模糊,降低信息冗余,通过改进THT(Top-Hat Transform)机制,设计了一种新的红外-可见光图像融合算法.首先,为了有效地提取源图像的特征和细节,在传统的高帽变换(Top-Hat Transform,THT)中引入了多尺度计算,分别在不同尺度上提取红外和可见光图像的不同亮(暗)特征区域;其次,通过多判别对比融合规则来综合两个图像中的同一尺度的特征区域,将所有尺度上的特征进行累加,获取特征图像;然后,通过最大尺度结构元素的开、闭运算得到源图像的平滑明亮、暗图像,在平滑图像上利用Gaussian模糊逻辑融合规则获取基础图像;最后,将提取的亮与暗特征图像导入基础图像中形成新图像,输出图像融合.实验表明:与当前流行的融合方法相比,本文方法具有更好的视觉质量和定量分析结果.
[Abstract]:In order to improve the fusion quality of infrared (IR) and visible (VI) images, maintain good visualization details and highlight IR targets, effectively reduce blur and information redundancy, the mechanism of THT (Top-Hat Transform) is improved. A new infrared-visible image fusion algorithm is designed. Firstly, in order to extract the features and details of the source image effectively, multi-scale computation is introduced into the traditional high-cap transform (Top-Hat Transform,THT) to extract different bright (dark) feature regions of infrared and visible images at different scales. Secondly, the feature regions of the same scale in the two images are synthesized by the multi-discriminant contrast fusion rule, and the features on all scales are accumulated to obtain the feature images. Then, the smooth bright and dark images of the source image are obtained by the opening and closing operation of the maximum scale structure element, and the basic image is obtained by using the Gaussian fuzzy logic fusion rule on the smooth image. Finally, the extracted bright and dark feature images are imported into the basic image to form a new image, and the output image is fused. The experimental results show that this method has better visual quality and quantitative analysis results than the popular fusion method.
【作者单位】：安阳工学院计算机科学与信息工程学院;
【基金】：国家自然科学基金项目(U1204613) 河南科技攻关计划重点项目(122102210138)
【分类号】：TP391.41

【相似文献】