Font Size: a A A

Research On The Application Of Image Fusion Technology In Process Tomography

Posted on:2019-03-16Degree:MasterType:Thesis
Country:ChinaCandidate:Z N SiFull Text:PDF
GTID:2428330548986589Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
There are many problems of distribution of multiphase flow in the process of modern industrial monitoring.The accurate detection of the distribution of phases is of great significance for the optimization and control of industrial processes.Therefore,the process tomography has developed rapidly in recent years.However,the limitations of the single modal imaging mechanism,the ill-posedness of the reconstruction equations and the real-time requirements of system applications place high demands on the hardware and the image reconstruction algorithms.This topic will focus on the principle of image fusion,meaning and algorithm to explore.By analyzing the imaging mechanism of Computerized tomography(CT),Electrical capacitance tomography(ECT)and the limitation of single mode,the image fusion technology is applied to image reconstruction of CT and ECT.The main work and results of the dissertation are as follows:(1)This paper proposes an improved logical filter fusion algorithm based on spatial domain to be used in CT/ECT image fusion,and uses the similarity measure of pixel values to perform OR operation.For the process tomography system,when the wavelet coefficients of CT and ECT reconstructed images differ greatly with different imaging mechanisms,this algorithm can improve the 0,1 information of the reconstructed gray image,which makes it close to the binary image of the ideal flow pattern.At the same time,the algorithm is simple and easy to implement,which is conducive to real-time monitoring of industrial sites.(2)Based on the advantages of D-S evidence theory in dealing with the uncertainty and fuzziness problems,this paper proposes to apply D-S evidence theory to CT/ECT image fusion.By using the membership Function represents the basic probability assignment function in evidence theory,to achieve "pixel-probability-pixel" conversion.The experimental results show that the proposed method can effectively improve the image reconstruction accuracy,and has less interference to the real-time performance of the system.
Keywords/Search Tags:CT, ECT, image fusion, improved logical filter algorithm, D-S evidence theory
PDF Full Text Request
Related items