Study On Local Reconstruction Algorithm For Fan Beam CT

Computed tomography (CT) imaging technology is the most widely used today, and isthe most practical of nondestructive testing technology. Fan beam scanning is simple, easy toimplement and control, and thus has been widely used. Place classic fan beam filter backprojection (FBP) algorithm is a global reconstruction algorithm, if we want to rebuild part ofthe image, you have to get the whole object of projection data, which will increase the X-rayradiation of the detected objects. Combined with the engineering often encounter big objectreconstruction, and the problem of high resolution reconstruction, fan beam localreconstruction problems is becoming more important.In this paper, two different kinds of fan beam local reconstruction algorithm are putforward, which is based on the theory of wavelet and PI line theory, now the main statementsare as follows:(1) To analyze the FBP algorithm of non-locality.(2)To study the theory of wavelet transform. According to the wavelet multiresolutionanalysis feature to study the fan beam local reconstruction algorithm which is based onwavelet. Its theory is based on certain scale function disappear of higher order moment ofcompactly supported by Hilbert transform, and it will get basic consistent support interval andthe original scale function. According to this theory, the wavelet coefficients and scalingfunctions are obtained directly by the projection data, which is to modify the Ramp filter, andthen place with FBP algorithm to solve it.(3) To analyze local reconstruction algorithm based on PI line, it is differential projectiondata, backprojection, and then for the limited Hilbert transform in the PI online, this method isdifferent from the traditional place of FBP algorithm, it is the backprojection filtration (BPF)algorithm. This algorithm only needs to be interested in regional area scan, for part of theprojection information can rebuild the desired image, it is a kind of accurate reconstruction algorithm.