摘要
异源的数据融合可以有效的综合各数据所包含的丰富的信息。针对变换域的融合方式使图像生硬不自然且计算复杂度高,介绍了一种利用加速引导滤波在空间域对红外图像和可见光图像的融合的方法。首先将红外图像进行滤波使目标高亮,然后将原红外图像作为加速引导滤波图对高亮的红外图像进行加速引导滤波,最后滤波结果直接与可见光图像融合,达到将异源图像融合的目的。该方法不但能够保留可见光图像的细节,而且能够凸显红外图像中的目标,为后续图像处理等手段提供更丰富完整的信息。四组对比实验表明该方法运算耗时少,融合效果较好。
The complex information can be effectively integrated through multi-source-data fusion method. Considering the inaccurate result and complex computationally via fusion method of the transform domain, in this paper, a method for fusing infrared and visible images with fast guided filtering was proposed. Firstly, the infrared images were processed through a filter to highlight the object. Then the previous images were analyzed by using original infrared images as the fast-guided filter through. Finally, the final filtering results were fused with the visible images to achieve image fusion. Not only the details of the visible image can be preserved, but also the object in the infrared image can be highlighted by using the proposed method. Comparing with other methods, the experiment results showed that the proposed method consumes less time and performs better. This result will surely provide rich and complete guidelines for the future research.
作者
陈洋
王世峰
都凯悦
王锐
CHEN Yang;WANG Shifeng;DU Kaiyue;WANG Rui(School of Optoelectronic Engineering,Changchun University of Science and Technology,Changchun 130022)
出处
《长春理工大学学报（自然科学版）》
2018年第6期11-15,共5页
Journal of Changchun University of Science and Technology(Natural Science Edition)
基金
吉林省科技厅自然科学基金项目(20150101047JC)
