Computed Tomography Images vs. Classical Radiographs
The radiography image grey values reflect both thickness and (radiographic) material density in the respective object area. So, thick areas appear darker than thin areas and lighter material appear brighter than heavier materials. In this way superposed features overlap in the radiograph. Typically, porosities look like bright spots while inclusion appear as darker spots. This different radiation intensity is caused by attenuation of the transmitting X-ray according to the Beer-Lambert law.
The grey values in the tomographic images represent the material at the respective three-dimensional position within the object under examination. Typically, denser material is displayed in brighter grey than less dense material. Iron for example will appear quite bright, aluminum medium, and air almost black.
The shape of the different parts of the object is directly visible and can be rendered by specific software thus providing a three-dimensional representation of the object under examination.
Beer-Lambert-Law
An X-ray beam of intensity I 0 transmitting through a workpiece and is attenuated according to the Beer-Lambert law. All object elements along the path of the beam contribute to the attenuation that results in an intensity distribution I(x, y). The contribution depends on the so-called attenuation coefficient μ of the material which describes the local attenuation.
Classical Radiograph
In the detector pixels the Intensity distribution of the X-ray beam is converted into numbers (pixel value), which are then displayed on the monitor as a distribution of grey values (X-ray image).
Computed Tomography
By taking X-ray images of the workpiece while simultaneously rotating the workpiece, information about the position of the various features is collected and retrieved by a computer-aided numerical process, the reconstruction. This process creates a three-dimensional volume (tomogram) consisting of voxels (volume elements) to each of which a voxel value is assigned, that is ideally proportional to the attenuation coefficient.
We can say
A digital X-ray image is a rectangle consisting of pixels, each of which has a pixel value that describes the intensity of the X-rays transmitted through the sample. This corresponds to thicknesses and material type.
A tomogram is a cube consisting of voxels each of which carries a voxel value that (ideally) represents the attenuation coefficient of the local material.