Dual Energy Single Slice Helical CT Reconstruction Algorithms For Luggage Security Inspection System

In recent years, public terrorism, drug smuggling and food safety affairs occurred frequently, which turns public security into an issue of worldwide concern and in turn, places a higher expectation to the security inspection. CT baggage scanning system is effective to detect the explosive and drug, and dual energy single-slice helical CT (SHCT) has been widely adopted. With its high scanning speed, compact design and low cost, single-slice helical CT will have a good market prospect. This paper will make a systematic study on the reconstruction algorithms of SHCT and focus on high pitch dual energy SHCT reconstruction algorithm and a correction method with dual energy DR image used on the CT image.This paper studies the scanning procedure and reconstruction method of the present SHCT, analyzes the problems of low resolution of Z direction and the helical artifact between layers during approximate reconstruction. In order to solve these problems, this paper proposes a new reconstruction algorithm for single-slice dual energy CT.1. This paper proposes the ART+3DTV reconstruction algorithm for single-slice helical CT. By studying the ART+TV algorithm used on limit-angle problem, this paper introduces 3DTV minimization limitation into the ART iteration, in order to solve the special limit-angle problem, in which each slice contains 36 degree projections. This method reduces the required data during reconstruction, accelerates the convergence of the iteration, increases the reconstruction slices to 5-10 times and reduces the artifact between slices.2. This paper proposes CT image correction method with DR image based on dual energy CT and DR scanning. This method eliminates the different effect of different systems by making the two data system geometrically compatible and using Compton-Photoelectric effect to resolute the dual energy image data, so as to make the two data system physically unified. After DR correction, the CT reconstruction result realizes higher accuracy. In addition, this algorithm needs huge calculation, so we need parallel algorithm to make it practical. And this paper uses GPU to accelerate the reconstruction procedure, and the acclerated algorithm can reconstruct the metarial information of a 256×256×256 baggage data in 46 seconds, which can completely meet the requirement of algorithm research and reach the demanded inspection throughput.