Research Of Counting System For Oil Casing Based On Image Processing

Petroleum casing pipe is a kind of steel pipe,which is used to support the well wall of oil and gas well.Each well uses several layers of casing according to different drilling depths and geological conditions.It belongs to single consumption materials which can't be reused.As a result,the consumption of the casing is very large.However,manual inventory not only requires high labor strength but also has a series of shortcomings,such as unstable precision,low efficiency,error prone and especially its incompatibility with the current automatic production and management.In current paper,several automatic counting methods based on image processing are put forward to solve the problem of low efficiency and low accuracy in manual stock taking.First of all,the result of the method based on gray threshold and edge detection depends largely on the quality of image binarization and have poor anti-interference ability.Secondly,the method based on template matching is used to identify the casing end face.Although the negative factors of the coincidence image formed by the noise and the placement position in the binary image on the counting accuracy can be neglected by the template method,it takes long time to shift the template in all directions which results in poor real-time.In addition,this method is sensitive to the size of the target.Finally,considering the obvious color information of the casing end face,the accuracy of the HSV color feature segmentation and the anti-interference ability of the illumination are fully utilized.The target region is segmented by the high and low thresholds of the HSV spatial chrominance channel,and then the mathematical morphology is introduced to overcome the interference in the area of cavity adhesion in order to provide a clear outline for the following counting.This method can overcome the difficulty of precisely detecting the edge of the casing area,so it is easier to achieve accurate counting.In the compiler environment of configuring the OpenCV2.4.4 in the VS2010,experiments were carried out with three proposed methods.The oil casing counting method based on HSV space and morphology is validated.Accurate identification results are obtained and the real-time requirements are met.So according to the algorithm,the software and hardware exploitation environment of handheld oil pipe counter is set up on the basis of S3C2440 development board and embedded Linux executive system.In addition,the facility application program of counting system is designed in Qt,and the automatic counting function of petroleum casing is preliminary realized.