Study And Implementation On The FDK Algorithm Of Three Dimension Cone-Beam Based On DSP

The explosives have become the important test item at the airport baggage inspection because of its hazards to the air visitors and the public safety. The explosives hidden in airport baggage have some characters such as small volume, variable shape, and the difficult measurement of its weight. The research on new detection techniques, its algorithms, and the related automation equipment shows great importance. At present, the CT(computed tomography) is the popular way for the airport baggage screening; amd it is detected by the density of explosives to get the accurate location. However, the volume of CT equipment in airport is often too large; further more, with the development of the aviation industry, the larger number of passenger baggage must require more high detection speed. The image reconstruction algorithm is the key point in the CT technology; and the FDK algorithm has played most important role. The disadvantage of this algorithm is the much time consumption;therefore decrease the detection of the CT equipment seriously.In recent years, the embedded DSP(digital signal processor) technology has become popular. It is usually applied in the computing of the complicated algorithm requiring high speed, and reducing the equipment volume. For the air status of the three-dimensional detection of explosives issue, the DSP solution is proposed in this paper. And the main work includes the fully study on the FDK algorithm of 3 Dimension Cone-Beam; combining the careful research on the limited space resources of DSP inner chipset to enhance the speed of the FDK algorithm from the hardware aspect; using the EDMA and USB2.0 protocol via BULK way to increase the data T/R(transmit/receive) speed in order to reduce the time cost in reconstruction step of the FDK algorithm; and it finally achieve the purpose of detecting speed CT equipment at the same time. In details, the related CT image reconstruction theory,the FBP algorithm in 2Dimensions since it is the basis of the FDK, and the FDK algorithm itself; and then the characters of TI TMS320C6416 T is firstly studied for the FDK application;and the follow-up contents are the the study on the optimization of the FDK algorithm itself using the spatial correlation which mainly including the transformation data type from the float point fixed point, optimization in cache computing and the EDMA for related date T/R;and the research on the USB2.0 protocol via BULK way to implement the data T/R between the USB device and the USB host, and the display of the reconstructed image. Experimental results show the feasibility of the method used in this paper and the speed improvement for the FDK algorithm computing via DSP after the guaranteed quality of reconstructed image.