Research and accuracy analysis of shaft parts size detection technology based on image processing Huang Jiexian Abstract: According to incomplete statistics, the annual output of shaft parts in China is about 654.38 billion pieces, of which about 70% need to be measured. At present, as far as the size detection of parts in many domestic manufacturing industries is concerned, the detection work still stays at the stage of manual sampling inspection of products by pure artificial vision or the combination of artificial vision with mechanical measuring tools and optical instruments [1]. Manual detection often has shortcomings such as low efficiency, poor reliability, low detection accuracy, high cost and easy to make mistakes. It is not suitable for the development of modern industrial enterprises. The size detection method based on image detection can not only avoid the shortcomings of manual detection, but also realize online, fast, accurate and non-contact automatic detection of machined parts. However, in the current research work of shaft parts detection based on CCD, the detection accuracy is still not high. And the detection data is not stable enough. Based on the development trend and practical application requirements of the discipline, this topic focuses on the high-precision detection technology of shaft parts size based on image processing on the basis of consulting a large number of documents and analyzing the CCD data acquisition system in the industrial field. The main work of this paper is as follows: (1) Prewitt operator is used to complete the initial location of image edge, and on this basis, the extreme value of the fitting curve is calculated by fitting the discrete value of image edge gray change with least square curve, and the accurate position of the edge is obtained. (2) In order to reduce the influence of interference on the measured values, the error data processing method is adopted to screen out the detection data with certain accuracy, and then these detection data are averaged to obtain stable detection data. (3) In order to solve the error problem caused by lens distortion in the high-precision detection of linear CCD, an error distortion correction model is proposed based on the known multi-dimensional axis parts, and the detection value is distorted to realize the theoretical distortion correction of the least square curve fitting error in the image processing of linear CCD.

Degree awarding unit: Guangdong University of Technology

Degree level: Master's degree

Year of degree award: 2008

Classification number: TP39 1.4 1

Doi: China HowNet: CDMD:2008. 36360.88638686667

Table of Contents: Summary 4-5 Summary 5-6 Table of Contents 6-9 Content 9- 12 Chapter 1 Introduction12-161.1Significance of research. 0. 1.2 Present situation and shortcomings of domestic measurement technology12-1.3 Significance of this research 13 1.2 Research status and development trend of shaft parts size detection technology based on image processing/kloc-0 Kloc-0/4654338 Hardware Design of Image Measurement System 16-222. 1 System Composition 162.2 Precision Mechanical Displacement Scanning Control System 16- 172.3 Linear CCD Camera17. 9-202.6 Design of clamping table 202.7 Computer and processing software 20-2 12.8 Overview of this chapter 2 1 -22 Chapter III Accurate Positioning of Axle Parts' Edge Based on Image Processing 22-343.438+0 Principle of Image Processing 22-253.438+0 Purpose of Digital Image Processing 2003.0000000006 Main Research Contents of Image Processing 22-243.63 Edge Position 3 1-333.4 Section 33-34 of this chapter. Kloc-0/ One-dimensional Normal Distribution 34-364.2 Regularity of Accidental Error 36-394.2. 1 Analysis of Accidental Error 37-394.2 Overview of this chapter 4 1-42 Chapter V Distortion Correction 42-495. 1 Generation of Distortion 496.2 Interface Design and Function Description 49 Kloc-0/-52 Chapter VII Experimental Results and Data Analysis 52-567.5438+0 Multistep Axis Measurement Results 52 8+0 Measurement Data 52-547. 1.2 Analysis of Measurement Results 547.2 Error and Precision Analysis 54-557.3 Summary and Prospect of this Chapter 55-56 56-58 References Welcome: Buying Network Cards.