Submitted to: Applied Engineering in Agriculture
Publication Type: Peer reviewed journal
Publication Acceptance Date: 8/1/2007
Publication Date: 10/1/2007
Citation: Haff, R.P. 2007. Real time correction of distortion in x-ray images of round objects and its application to agricultural commodities. Applied Engineering in Agriculture. 51(1):341-349 Interpretive Summary: Many food products are either spherical (apples, oranges, etc.) or cylindrical (bottles, cans, jars) in shape. This causes problems in x-ray images because the x-ray power required to penetrate the thick part of the sample tends to wash away the thinner parts. The result is an image that appears out of focus with detail missing along the edges. In this study, a technique to construct “lenses” which correct this deficiency in the x-ray images is derived and demonstrated. The technique can be applied either to x-ray images of stationary samples, common in quality control sampling, or moving samples, common in mass inspection in processing plants. The shape and material characteristics of the required lenses are derived, and constructed and tested for the case of an apple, a lime, and an orange. The resulting images demonstrate that the lenses produce higher quality focused x-ray images.
Technical Abstract: A technique to correct the deficiencies in x-ray images of round objects that are a consequence of the geometry of the sample is derived. The technique is demonstrated in x-ray images of an apple, an orange, and a lime, but applies equally well to any sample with a spherical or cylindrical shape. The methods described, which involve the use of molded trays (“lenses”) to equalize the x-ray absorption across the sample, are inexpensive and easy to implement. The lenses demonstrated here are useful only for sampling type of inspection. Application of the method for real linescan x-ray systems capable of high speed sorting is derived and discussed.