Abstract: This paper elaborates the basic knowledge of image fusion，and improves the traditional algorithm of multi-focus images fusion based on spatial frequency.It specifically refers that calculating the window’s spatial frequency of every pixel as the pixel’s value and then processing the location of the pixels according to the value of the spatial frequency and finally generating a fusion image all of whose objects of the same scene are clear.In order to verify the effectiveness of the approach，a simulation test is designed.The result shows that the algorithm is better than traditional algorithm in multi-focus images fusion.

Key words: image fusion, multi-focus images, spatial frequency, quality assessment

摘要： 主要论述了图像融合的基本知识，对多聚焦图像传统的基于空间频率的融合算法进行了改进。分别计算多聚集图像对应位置上每个像素所在窗口的空间频率，然后根据空间频率的大小对该位置像素进行处理，最后生成同一场景中所有物体都清晰的融合结果图像。通过仿真实验验证了算法的有效性，结果表明，该算法在多聚焦图像融合上要优于传统算法。

关键词: 图像融合, 多聚焦图像, 空间频率, 质量评价