Material Discrimination based on Atomic Number of Scanned Items using Dual Energy X-Ray Images

An agreement was signed between M/s. Krystalvision Image Systems Pvt. Ltd. and CSIR-CEERI on 4th November 2015 for undertaking the above project. The project was commenced on 4th November 2015 and its duration was for seven months.

This project was aimed at developing material discrimination technique based on an atomic number and density of scanned items using dual energy X-ray images. M/s. Krystalvision would provide dual energy X-ray image data and images of 10 to 20 materials taken from their X-ray imaging system for the analysis and determination of atomic number & density. CSIR-CEERI would conduct a feasibility study on the problem of determining the atomic number and density from the given dual energy X-ray image data and propose a solution by formulating necessary theoretical equations. Further, CSIR-CEERI would analyse the X-ray image data and develop a dual energy image processing technique for the measurement of atomic number and density of selected materials. The developed technique would be integrated with the dual energy X-ray system and tested using test piece provided by M/s Krystalvision. On successful testing, the software could be used to identify selected sample materials provided by M/s. Krystalvision. Also, the company is interested in taking the developed knowhow for commercialisation.

  • A list of elements and compounds with different material discrimination colour label based on atomic number was prepared and sent to M/s Krystalvision for image capture. An image databank was generated for various materials by conducting X-ray image experiments using Krystalvision dual energy system at 140 KV @700 μA.
  • Direct reconstruction dual energy image processing algorithm was developed using a set of dual energy X-ray images of different materials given by M/s Krystalvision. Ten numbers of useful images were chosen for conducting the experiment. The image pixel value variation in the low energy (LE) and high energy (HE) was studied for each material using X-View software tool (Ver. 5.3). The pixel values along with atomic number were tabulated for each energy level using the images provided. The relevant mathematical model for computing the atomic number and density was derived and the same was implemented in the algorithm to display the atomic number image with a color label (3 colours or 6 colours).
  • A technical feasibility report titled ”Material discrimination using dual-energy X-ray imaging” was prepared and sent to M/s Krystalvision, Pune.
  • The CSIR-CEERI project team has demonstrated and explained the first level atomic number determination algorithm to M/s Krystalvision team. The software demonstration of atomic number image and attenuation image for various materials obtained by CSIR-CEERI-software was shown to them. The first level image output of Krystalvision Combined Test Piece (CTP) is shown in the following Fig.112.