摘要： 研究基于视觉伺服的自主微装配方法，为实现深度信息估计，提出一种基于马尔可夫随机场的散焦特征参数模型，将散焦特征深度信息的估计转化为能量函数的优化问题，采用Broyden 图像雅可比矩阵实现视觉伺服控制。通过微夹钳自动定位与夹取实验，验证了该方法的有效性。

关键词: 深度估计, Broyden方法, 图像雅可比矩阵, 视觉伺服, 微装配

Abstract: This paper studies the independent micro-assembly method based on visual servo. In order to realize the depth information estimation, it presents a blur parameter model of defocus image based on Markov random field for performing an effective and reliable automatic micro- assembly, which converts the problem of depth estimation into optimization problem. An improved Broyden method for estimating the image Jacobian matrix online is presented as the visual servo method. It designs the experiment of microgripper fix position and clip. Experimental results confirm that the proposed methods are effective.

Key words: depth estimation, Broyden method, image Jacobian matrix, vision servo micro-assembly

中图分类号: 

• TP242