Reseash To Tubing Collar Identify Problem Of Workover Based On Digital Image Processing

In the oil and gas exploration and production process,workover is an important factor.As the workover task has become increasingly arduous,labor intensity is very high.Therefore, we need to improve the efficiency of workover,reduce labor intensity of workers,it requires a higher degree of automation in workover operations to resolve or alleviate the above problems.Thus,the automatic detect to tubing collar as an important goal.Digital image processing and pattern recognition technology is that Involved in artificial intelligence, neurobiology,computer vision,image processing,mathematics and other fields of cross-disciplinary.As the development of technology,a large number of used space and aviation technology,industrial and engineering,military police and so on.This issue belongs to the oil industry applications.This paper combines the digital image processing technology, based on the analyzing the existing image pre-processing methods and moving targets recognition algorithm,tubing collar of single-frame images, and video image processing and algorithm analysis,and develop a coupling automated image recognition software,achieved some results.Major research work includes the following three aspects:First of all,Study of existing image pre-processing methods,Including image enhancement, image segmentation.Application of different algorithms of various methods on the collar single-frame image,Comparative analysis treatment effect of various algorithms,Select the best suitable for collar image pre-processing algorithm of workover environment.Secondly, use the Simulink in the MATLAB Component Video and Image Processing Blockset blocks, designed several tubing collar video image processing and marking system.through simulation test the processing effect of the video image algorithms, lay the foundation for the development of automatic identification software systems.Finally, for the tubing collar in which the specific environment, research and analysis for recognition algorithm of tubing collar image. based on VC++ software development platform and OpenCV(Open Computer Vision),Design and development of a tubing collar image automated image recognition system.Experimental results show that the background in a simple case, Has a good recognition effect.