Methods for reconstructing an unknown object in a scanned image

A method for assigning attributes to an unknown object comprising:

• scanning the unknown object within a container and at least partially overlapping with a background object within an x-ray scanning device, the x-ray scanning device emitting x-rays from at least two sources which pass through the object of interest and the background object, the x-rays being detected by at least one array of detectors to provide a plurality of dual-energy attenuation images each having dual-energy attenuation information representing the container and an overlap region wherein the background object and the unknown object and container overlap;
• decomposing each of the dual-energy attenuation images into reference material equivalent path length images;
• removing the reference material equivalent path lengths representing the background object from the reference material equivalent path length images to provide reference material equivalent path lengths representing the unknown object and the container;
• converting the reference material equivalent path lengths representing the unknown object into unknown object path lengths multiplied by a predetermined scaling factor;
• determining the effective atomic number for each pixel representing the unknown object and the container;
• determining the mass thickness for each pixel representing the unknown object and the container, the mass thickness being equivalent to the unknown object path lengths multiplied by the scaling factor;
• identifying each first source-detector pair line defined by a first x-ray extending between a first one of the at least two sources and one detector of the array of detectors in a first one of the plurality of dual-energy attenuation images on which lies one scaled unknown object path length;
• identifying each second source-detector pair line defined by a second x-ray extending between a second one of the at least two sources and one detector of the array of detectors in a second one of the plurality of dual-energy attenuation images on which lies one other scaled unknown object path length, the second one of the plurality of dual-energy attenuation images generated contemporaneously with the first one of the plurality of dual-energy attenuation images;
• joining the extremities of each of the scaled unknown object path lengths to provide a contour of the unknown object;
• iteratively matching the contour of the unknown object of each of the first and the second one of the plurality of dual-energy attenuation images to reduce the scaling factor of the scaled unknown object path lengths and to provide unknown object path lengths;
• defining the contour of the unknown object as an inner contour of the container;
• identifying third source-detector pair lines defined by third x-rays extending between each source and one detector of the array which intersect with the container at only one point of intersection in each of the first and second one of the plurality of dual-energy attenuation images and delimit an outer bound of the container as the pixels within the third source-detector lines;
• interpolating the outer bound of the container extending between the one point of intersection of each third source-detector pair line to define an outer contour of the container;
• determining path lengths representing the container as path lengths which extend between the inner contour of the container and the outer contour of the container;
• and,
• determining a density of the unknown object and an effective atomic number of the unknown object.