一种基于多项式的图像离焦建模方法研究

本文针对显微视觉的图像恢复与3D 重建问题,从显微光学成像的离焦机理和基于点扩散函数的图像模糊化描述出发,通过模糊测度算子分析序列显微图像的离焦分布规律,提出了一种用于构建较为精准离焦模型的方法。该模型采用混合参数多项式结构,与传统高斯模型相比,可以更接近真实离焦过程,这为较为精确的光学显微图像恢复和3D 重构提供了新的技术途径。

Based on the basic defocus principle of the Micro- Optics imaging and the image blur description by PSF(point spread function),this paper proposes a new method to obtain a more accurate defocus model. Through analyzing the disciplinarian of defocus distribution in the image sequences obtained by microscope, we construct a defocus degradation model described by the mix parameter polynomial.Experimental results show that this defocus degradation model is more approximate to the actual defocus process than other exis- ted models.That can give a new tech for image restoration and 3D recovering in micro imaging.

国家自然科学基金(60675060);;辽宁省人才基金(3050002)资助项目