Research On Imaging Distortion Mechanism And Image Correction Method Of Ultra-Wide-Angle Lens Wellbore

Ultra-wide-angle lens is widely used in the field of oil and gas wells due to its short focal length and large viewing angle.The downhole video images obtained by the ultra-wide-angle lens can capture more abundant image information in the wellbore within an effective range.However,obtaining a clear image with a large viewing angle will produce large image distortion,which will affect the analysis results of logging data.Therefore,it is very important to correct the distorted images before analyzing and processing the data.The paper mainly includes the following four aspects : camera model research,ultra-wide-angle camera calibration,image correction and expansion,and algorithm software implementation.Aiming at the imaging problem of downhole ultra-wide-angle camera,the imaging distortion mechanism is studied according to the imaging characteristics of the camera.The structural characteristics of the camera and the actual observation angle of the camera are analyzed,and the causes of wellbore distortion are explained.Based on the research of two camera models-UCM and EUCM models,a camera model suitable for ultra-wide angle-threedimensional perspective model is constructed.The problem solves that the traditional model cannot simulate the projection process of ultra-wide angle camera.The camera coefficient is obtained by camera calibration.The image captured by the ultrawide-angle camera produces great distortion,which makes the traditional linear calibration method no longer applicable.Aiming at the application environment of ultra-wide angle,the model parameter calibration method is adopted.Through the real-time calibration of casing coupling and pipe diameter data,the optimization method is used to adjust the parameters to match the model.The calibration process is divided into ideal wellbore and actual calibration wellbore calibration experiments.Firstly,an ideal checkerboard wellbore model is established to obtain the camera coefficients and verify the feasibility of the calibration algorithm.Then the hardware platform is built for camera calibration experiments to solve the camera coefficients of the real camera.Aiming at the problem that the global projection correction method is difficult to apply to the three-dimensional environment of the wellbore,a three-dimensional grid matching image correction algorithm based on perspective model is proposed.The algorithm establishes a threedimensional wellbore model based on the perspective model imaging principle,and uses the specific information of the input image to perform three-dimensional mesh matching between the model and the distorted image.By selecting the effective area of the target segment image,the pixel value is obtained by traversing the ring area with the backward mapping method,and the pixel points of the ring area are refilled into the rectangular expanded image to realize the correction of the distorted image.The image clarity and correction effect after correction are better than those processed by the longitude coordinate correction algorithm.Modularize the algorithm proposed in this paper,build a software platform,design image correction software,and perform software testing.The test results show that the software has stable performance and good correction effect on distorted ultra-wide-angle images.The software is applied to the processing of oil and gas well logging video,and the measurement error of the corrected image is within ± 1 mm,which has a good application prospect in the field of oil and gas wells.