摘要
针对多聚焦图像融合,从人眼视觉对比度以及多聚焦图像清晰与模糊像素之间的边界点出发,提出一种基于有理三次样条的图象融合算法。将多聚焦图像划成三个区域:清晰区域、模糊区域和边界区域。在边界区域,利用有理三次样条函数构造权函数,并以边界区域像素为中心,用清晰区域的像素和模糊区域像素做加权组合,计算边界区域像素的灰度值。经实验验证,提出的图像融合方法具有较强的灵活性和可靠性,具有较好的融合效果。
To address the multi-focus image fusion issue, this paper proposes an image fusion algorithm which is based on rational cubic spline from the perspective of human visual contrast and boundary point between clear and blur pixels of multi-focus image. Multi-focus image is divided into three regions ： the clear region, the fuzzy region and the boundary region. In boundary region, the rational cubic spline function is employed to construct weight function. Moreover, the pixels of boundary region are taken as the centre, the pixels in clear and fuzzy regions are used for making the weighted combination, and the gray value of pixels of boundary region is then calculated. Proven by the experiments, the image fusion method proposed here has strong flexibility, reliability, and good fusion effect.
出处
《计算机应用与软件》
CSCD
北大核心
2013年第1期151-154,共4页
Computer Applications and Software
基金
安徽高校省级自然科学研究项目(kj2011z285)
滁州学院科研基金项目(2010kj008B)
关键词
图像融合
多聚焦图像
有理三次样条
权函数
清晰度
Image fusion Multi-focus image Rational cubic spline Weight function Clarity
