Algorithm of high-dynamic fusion image gray characterization based on variable energy
- 1. Key Laboratory of Instrumentation Science and Dynamic Measurement of Ministry of Education, Shanxi Key Laboratory of Signal Capturing and Processing, National Key Laboratory for Electronic Measurement Technology, North University of China, Taiyuan 030051, China;
- 2. Key Laboratory of Molecular Imaging, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China
- Received Date:
- Accepted Date:
- Available Online:
Project supported by the National Natural Science Foundation of China (Grant Nos. 61227003, 61171179, 61302159), the Natural Science Foundation of Shanxi Province, China (Grant No. 2012021011-2), the Specialized Research Fund for the Doctoral Program of Higher Education, China (Grant No. 20121420110006), the Shanxi Scholarship Council of China (Grant No. 2013-083), and the Top Science and Technology Innovation Teams of Higher Learning Institutions of Shanxi, China.
Abstract: X-ray imaging based on variable energy can expand the dynamic range of the imaging system and perfectly show the structure information of the detection objects, by acquiring and fusing the image sequences. However, the fusion method is ordinarily based on image quality optimization, and neglects the gray mapping accuracy of the actual high dynamic imaging. It cannot guarantee the physical matching between the image information and actual structure information. Therefore, in this paper we propose an X-ray image gray characterization algorithm of high dynamic fusion based on variable energy. First, take a standard wedge block as test object, and use the fusion image of low dynamic image sequences as input data. The output data are the actual high dynamic image. Then establish the X-ray imaging gray characterization model by neural network training. At the same time, because the attenuation coefficients of different heterogeneous materials are different, a modified model of physical characterization is established to achieve a correct characterization of real object. Finally, experiments by 12 bit and 16 bit imaging systems acquire the variable voltage image sequences using 12 bit detector. After image fusion, image mapping and gray level correction, the output image not only achieves the same effect of 16 bit detector, but also satisfies the gray relation. Also this method can effectively expand the dynamic range of the imaging system.