Font Size: a A A

Research On Flexible Recognition System For Robot Based On Binocular Vision

Posted on:2013-08-26Degree:MasterType:Thesis
Country:ChinaCandidate:L C GuFull Text:PDF
GTID:2248330377460244Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
In recent years, the automotive industry has been developing rapidly.Universal, efficient and stable system is becoming more important to handle theincreasingly complicated processes. The whole automotive process including thestamping, welding, coating and final assembly processes is intricate. The robotflexible recognition system can be integrated in body fabrication processes toenhance the flexibility. This paper adopts flexible robot recognition system basedon binocular vision, and mainly researches on Eye-to-Hand system. The paper alsoincorporates it into designing the coupling between the several deferent processesand the Eye-to-Hand system and finally constitute an efficient system with highflexibility, high accuracy.Firstly, this paper describes the importance of robot recognition technology inthe automotive industry. Based on analyzing the recent research status and futuredevelopments of industrial robots in the world, discussion of the recognitiontechnology, investigating on vision technique and combining the four process ofautomobile manufacturing, we find that the flexible robot recognition system basedon binocular vision has nicer practical applied foreground.Secondly, the paper described the recognition subsystems, grasped theconnection between them, analyzed internal and external system and identifiedinput-output sequential control system with the variation of fuzzy technology.Then, it mainly designed vision system based on some theories and algorithms,constructed the internal framework of vision system, and analyzed operationalprocess. At last, we can find a visual theoretical model that meet the recognitionsystem very well. According to the visual theoretical model, we analyzed thereasons for image degradations and methods of calibration of camera, and formedone method of improved visual calibration. A further analysis on the theoreticalmodel of binocular vision system have had been made. We also got a relatedalgorithm that provided this system with some theoretical support.At last, a mathematical model of this whole system has been constructed. Byconstructing and analyzing this model, we got the spatial relationships between the subsystems. Then, we focused on designing coupling network of this system,communicating with each process lines, meeting the requirements and so on. Thissystem has eventually been used in the press line to analyze and study therationality, versatility and other performance of this system. Finally, acomprehensive report has been made.
Keywords/Search Tags:Binocular vision, Industrial robotics, Mathematical modeling, Recognition technology, Press line
PDF Full Text Request
Related items