Abstract: This paper proposes a novel multi-focus image fusion method based on point spread functions (PSF). It uses Gaussian functions to convolve with multi-focused images repeatedly. Compared with approximation error images, it gets the focus and defocus regions, and segments the regions through morphological method. It uses choose-max fusion rules to get the all-in-focus image. Experimental results show that the image is distinctly segmented into multi-regions, the image edge is legible, and the fusion results are satisfactory.

Key words: point spread functions(PSF), error image, image fusion, morphological segment

摘要： 提出了一种新的基于点扩散函数的多聚焦图像融合算法，采用高斯函数对多聚焦图像进行多次卷积，通过与原近似误差图像的比较，得到聚焦清晰和模糊区域，采用形态学方法将其分割出来，并根据最大值融合规则对其进行了融合。试验结果表明，该算法可将图像分割成多个聚焦区域，边缘清晰、融合效果良好。

关键词: 点扩散函数, 误差图像, 图像融合, 形态学分割

CLC Number: 

• TP18