中文摘要：

为了提高工业生产的柔性和自动化控制程度，在传统的零件装配系统中引入视觉注意思想和协调控制方式。使用基于直方图对比的方法生成显著图，用固定阈值优化法进行二值化，利用最小外接矩形抽取出零件，通过特征提取和匹配识别出零件，用于指导工业机器人抓取，结合两台工业机器人主从式协调控制方式，完成零件装配任务。在自主开发的平台上进行实验，结果表明了所提出的方法能够快速地识别和定位出目标零件并且在仿真平台上成功地模拟了零件协调装配的整个过程。

英文摘要：

In order to improve the flexibility and the degree of automatic control in industrial production, visual attention idea and coordination control mode was brought into the traditional system of parts assembly. Firstly, saliency map was produced by the method of Histogram-based Contrast; secondly, parts was highlighted by minimum enclosing rectangle after using fixed threshold optimization method for binarization; and then, the parts was identified by feature extraction and feature matching to guide industrial robots to grab; lastly, combining coordination control mode of master-slave between 2 industrial robots, the task of part assembly was completed. Simulation experiments were conducted on the self--development platform. Experimental results show that the presented methods con identify and locate the target parts quickly and succeed to simulate the entire process of assembling parts concertedly.