Title :
The Depth Information Estimation of Microscope Defocus Image Based-on Markov Random Field
Author :
Zeng, Xiangjin ; Huang, Xinhan ; Wang, Min ; Li, Peng
Author_Institution :
Dept. of Control Sci. & Eng., Key Lab. of Image Process. & Intell. control, Wuhan
Abstract :
For depth information estimation of microscope defocus image, a blur parameter model of defocus image based on Markov random field has been present. It converts problem of depth estimation into optimization problem. An improved iterated conditional modes algorithm has been applied to complete optimization problem, which the select of initial point employed least squares estimate algorithm prevents that the result gets into local optimization. The experiments and simulations prove that the model and algorithm is efficiency.
Keywords :
Markov processes; image processing; iterative methods; least squares approximations; microscopes; optimisation; Markov random field; blur parameter model; depth information estimation; improved iterated conditional modes algorithm; least squares estimate algorithm; microscope defocus image; optimization problem; Cameras; Charge coupled devices; Charge-coupled image sensors; Image converters; Least squares approximation; Lenses; Markov random fields; Microscopy; Orbital robotics; Robot vision systems; Iterated Conditional Modes; Markov random field; depth estimation; microscope vision; the blur model;
Conference_Titel :
Robotics, Automation and Mechatronics, 2008 IEEE Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4244-1675-2
Electronic_ISBN :
978-1-4244-1676-9
DOI :
10.1109/RAMECH.2008.4681326