Research Of Moving Workpiece's Inspection And Measurement

According to the requirement of chromatic dislocation inspection application in printing industry,this paper combines CCD measurement technology,image analysis technology with pattern recognition theory. Based on these theories,the author designed On-line chromatic Bar Code Measurement System that uses CCD as image sensor and corrects error using neural network. The thesis presents the compositions of this measurement system,discusses relevant theory and design of these compositions. Through analyzing in theory and experimenting,the factors that affect measurement precision have been discussed. Meanwhile,corresponding resolutions are provided. On the basis of different characters of static and dynamic measurement,the author corrects error for the system using linear fitting and neural network,respectively,thus achieving satisfactory results. Finally,these research achievements are applied in the system's software design.This thesis introduces the hardware composition and measurement theory of general CCD data-acquisition system by referring to lots of documents and dissecting CCD data-acquisition systems used in the industry field. At the same time,the developing method of software's main modules is given in this paper. The system has obtained following effect:when the velocity of bar code is lower than 0.8m/s,its measurement accuracy reaches 0.1mm over 10cm range. Therefore,it satisfied basic requirement of industry on-line measurement.Research work and achievements may contribute to application and researching of pattern recognition and image analysis in metrology and measurement,and have certain practical value and economic benefits.