基于相对模糊和边缘强化的散焦求深度方法

摘 要: 研究了由散焦图像恢复物体深度的问题,改进了Favaro等人提出的基于迭代求解热扩散方程的方法.将两幅散焦图像按照相对模糊程度的不同,划分为两个区域,证明了这两个区域的边界和图像的边缘有很高的一致性,通过强化图像的边缘,使这两个区域的边界在迭代求解热扩散方程的过程中更容易确定.模拟实验与真实实验都验证了算法的有效性.

Depth from Defocus Based on Relative Blur and Edge Enhancement

Abstract: The issue of depth from defocus (DFD) is studied and Favaro’s iteration method based on the thermal diffusion equations is improved in this paper. According to the relative blur, the defocus images are divided into two regions respectively. It is proved that there is a very high consistency between the border of the two regions and the image edge. By strengthening the edge of the image, it is easier to identify the border of the two regions during solving the thermal diffusion equation by iteration. Simulation experiment and real image experiment both verify the effectiveness of the method.