The Developing Of Raw Cotton Foreign Fibre Online Detecting System

Machine vision has been the outstnading field in industrialautomation after deeades'development.Image measurement and image recognition are playing more and more important part in online inspection system in industry.Cotton contains noncotton fibre has an important effect to the cotton textile,Most important to the quality of the production.So it can affect the economic benefit of the enterprise and bring disadvantage influence in improving it's compete ability in market economy simultancity.Therefore measuring and eliminating the isomerism fibre in cotton on line has a very practice and very important senseon improving on the technologic methods for elementary processing cotton and increasing the quality of the finished product of cotton textile and reducing the cost.So it has a very good market value and development foreground.This paper discuss about foreign fibers detecting technology and foreign fibers detecting system,and especially the algorithm.Based on the strict demonstration and analyses,use the three colored dimensions mathematical module of the cotton constructed in the model of RGB color,the image processing algorithm of least square method using segmented threshold is put forward,this method solves the influence of the camera nonlinear to the efficiency of picking system effectively.Combined the method of nonlinear double threshold analyses with the image processing algorithm based on differentiation can improve the image processing efficiency,and also make sure the reflect made by the program when system acquires the image data is real-time and accurate.This paper has realized two aspects breakthrough in white foreign fibers and hair detection.The whole system can meet the commands of this industry.Firstly,this paper introduce the research background and the development of the foreign fibers picking system,secondly,the author did some research on the arithmetic of picking impurities out based on the technique of image processing.For the third,the author applied the arithmetic to the system and had a satisfied result during experiments. After trials on the spot for several times,it is concluded that the system has the traits of picking quickly and effectively.For the third,the author found the way to build up the hardware platform to minimizing the disadvantage,especially did quite a lot of experiments on choosing lamps to get a better input signal.The software platform of the system is Visual C++6.0,and the system can successfully check and get rid of impurities out from cotton on the on-line,real time,non-touched,dynamically condition, using present technique of image processing fully,based on image gathering card,CCD camera,Mcu etc.