Research On Precise Measurement Method Of Fuel Rod Diameter Based On Image Processing

The pressurized water reactor core AFA fuel assembly is the most important component used in Daya Bay Nuclear Power Station.For fuel assemblies,various phenomena such as cracks,wear,deformation,and thickening of the oxide layer will occur in the operation of the core.When the problem is serious,it will affect the integrity of the fuel assembly and the safe operation of the reactor.Real-time fuel rod inspection is required to ensure the safe operation of fuel rod assemblies.This paper designs a detection system that can measure the diameter of the fuel rod in real time based on image processing technology.By detecting the fuel rod in real time,the edge information is obtained and the diameter is obtained to determine the thinning or thickening of the fuel rod,making the fuel assembly more effective Put into operation.In this paper,a precision measurement system for fuel rod diameter based on image processing is designed.First,based on the analysis of the actual fuel rod collection environment,combined with the camera,lens,light source,calibration plate and fuel rod samples,a fuel rod hardware measurement system was designed and built.Complete the design of the camera calibration algorithm,use a standard checkerboard to calibrate the industrial cameras used,and use Zhang Zhengyou's calibration method to obtain camera parameters and achieve parameter optimization and camera distortion correction.Then,the noise contained in the actual fuel rod image was analyzed.Aiming at the noise existing in the image,different algorithms are used for denoising,but they all have the disadvantages of blurred edges and unclean noise points.Therefore,this paper proposes an improved median filtering combined with bilateral filtering algorithm.This method removes the noise better and retains the edges better.Because the overall grayscale of the acquired image is dark,the multi-scale Retinex algorithm is used to enhance the overall illumination of the image,and the multi-scale Retinex algorithm and the Laplace operator are used to enhance the details of the image.Aiming at the problem of tilting the fuel rod during shooting,the Hough transform method is proposed to complete the tilt correction of the fuel rod image.On this basis,after a variety of edge detection algorithms are compared on the fuel rod image,the Canny operator is used to extract the edge points of the fuel rod image with enhanced lighting.A variety of sub-pixel positioning algorithms are used to precisely locate the extracted edge points.This paper mainly improves the Gaussian curve fitting sub-pixel algorithm,and compared with other algorithms,the improved sub-pixel algorithm has higher stability and accuracy.At the same time,the sub-pixel algorithm in this paper is applied to the checkerboard corner point extraction,and the image pixel equivalent is obtained by combining the physical dimensions between checkerboard corner points.Finally,the software part of the fuel rod diameter ranging system is designed in this paper.From the aspects of camera acquisition and storage,parameter adjustment,calibration,image processing and size measurement modules to achieve the precise measurement function of the fuel rod.MFC software measurement interface is designed by using C ++ and Opencv open source library.Then the fuel rod diameter is precisely measured,and the pixel equivalent and the edge of the fuel rod are used to obtain the pixel-level and sub-pixel-level diameters,respectively.After comparison,it can be clearly obtained that the result of the sub-pixel edge is significantly better than the pixel-level edge.At the same time,the edge point can be selected in real time to obtain the diameter using the mouse drawing method.The fuel rod diameter precision measurement system designed in this paper measures the fuel rod samples at different distances,and the average error of the results is 9.88?m,which basically meets the accuracy requirements of the enterprise.