Dissertation > Excellent graduate degree dissertation topics show

Research on Inspection and Recognition of PCB Digital Image

Author: GaoXiangLing
Tutor: LinJianHui
School: Southwest Jiaotong University
Course: Measuring Technology and Instruments
Keywords: PCB Computer image processing Pattern Recognition
CLC: TP391.41
Type: Master's thesis
Year: 2002
Downloads: 765
Quote: 30
Read: Download Dissertation

Abstract


Early treatment is mainly used to improve the image quality of the image, with computer technology, image acquisition technology, the rapid development of image processing has become increasingly widely used to solve many fields of science and engineering problems, such as medical image analysis, fingerprint identification , product quality inspection. Since computer image processing has many unique advantages, which is developing very rapidly, becoming a research hotspot of modern technology. In the electronics industry, PCB are the main components of a variety of electronic products, PCB may cause defects on the printed circuit board or even the entire piece of equipment obsolescence, so PCB PCB manufacturers inspection and testing is an essential part of quality control. With the improvement of production technology, PCB production process is becoming increasingly complex devices and a large number of ultra fine traces are used to make the PCB quality inspection to become a very difficult task, it is difficult to automate. Traditional electrical measuring method and artificial visual method can not meet the production needs. The use of image processing and image recognition technology for PCB analysis in order to identify the defects on the PCB is a real-time, non-destructive, high precision and excellent low-cost method. Currently, the market has finished equipment sales, often referred to as automated optical inspection systems (AOI), but these devices are mostly foreign-developed, expensive, domestic still in its infancy; addition, the existing AOI system remains high The false positive rate. This shows that the PCB image processing and pattern recognition methods still need further study. Based on the above points, the thesis in digital image processing and pattern recognition based on the theory of the PCB image acquisition, processing, identification methods and difficulties in-depth discussion, and to use Borland C Builder6.0 prepared a set of PCB image detection and recognition software defects on printed board types and quantities of automatic detection and identification. PCB image detection and recognition system consists of four parts: image acquisition, image preprocessing, feature extraction and image recognition. In the image preprocessing, this paper analyzes the commonly used method of image enhancement and restoration of the applicability of the PCB image for PCB dim image, contrast is not strong and so gives a suitable PCB image feature approach. Image binarization, using bimodal and maximum variance threshold method for threshold, and achieved good results. In image recognition, this paper analyzes the commonly used pattern recognition methods, according to the classification tree and sequential probability ratio test method devised the idea of ??a unique PCB defect pattern recognition methods; and gives the defect modes for a variety of PCB feature selection and extraction methods. Meanwhile, the paper also right image acquisition system design and acquisition of the difficulties, we present a solution. In software development is completed, a large number of PCB image into the testing, after comparison test has proven to be effective in identifying various defects on printed boards, has received good results. Thesis design is successful, reaching the intended effect, research has good application value.

Related Dissertations

  1. The Classification of High Dimsnsion Flew Field Based on Manifold Learning,V231.3
  2. The Research on Action Mechanism of Additive Used in Through-Hole Copper Plating and Its Selection,TQ153.14
  3. Research on Text Classification Based on Biomimetic Pattern Recongnition,TP391.1
  4. Research on Identification System of Cashmere and Wool Fiber,TS101.921
  5. Intrusion detection based on the ultrasonic echo envelope in the military security patrols,E919
  6. Research on Key Technologies of Turning Expert System Based on SE300,TP182
  7. Improved Mathematical Model of Spatial Information Processing Based on Fuzzy Technology,O159
  8. The EMG Pattern Recognition System Based on Generalized Dynamic Fuzzy Neural Network,TP183
  9. Design and Implementation of System of Tables with Handwritten Content,TP391.41
  10. A Study on the Value Indictors System of Cultural Heritages in Henan,G122
  11. Control object based on a variety of new chips and chip Experimental Equipment,TP368.12
  12. Action surface EMG nonlinear characteristic of,TH772
  13. Based on RBF artificial neural network in the application of PCB drilling process,TN405
  14. GPRS-based wireless nodes tie breaker condition monitoring design and implementation,TN929.5
  15. CUDA-based video fire detection system,TP391.41
  16. Dollar bill image detection of critical information and number identification,TP391.41
  17. SVM Based on SIFT and scene classification,TP391.41
  18. Seamless integration of automatic detection of ultrasonic vortex systems research and development,TH878.2
  19. Based on DCE-MRI of computer-aided diagnosis of benign and malignant breast lesions studied,TP391.7
  20. Based on LabVIEW typical communication signal modulation , parameter identification and simulation experiment,TP274
  21. Smartphone dimensional code recognition system design and implementation,TN929.53

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile