Image feature based detection of agricultural quarantine materials in X-ray images

The US Department of Agriculture uses many methods to screen incoming passenger baggage for agricultural quarantined materials (e.g. fresh fruits and meat). X-ray inspection is relied upon because it is a fast and non-invasive inspection method. This paper investigates a new image feature based approach to training for X-ray inspection that instructs subjects to search for particular patterns and textures in the X-ray image, rather than specific objects (e.g. an apple). The image feature based approach is found to have demonstrable potential as a training program and provides justification for further research into implementation in a computer pattern recognition program.