Application of dual-energy x-ray techniques for automated food container inspection

Abstract

Manufacturing for plastic food containers often results in small metal particles getting into the containers during the production process. Metal detectors are usually not sensitive enough to detect these metal particles (0.5 mm or lesser), especially when the containers are stacked in large sealed shipping packages; X-ray inspection of these packages provides a viable alternative. This paper presents the results of an investigation into dual-energy X-ray techniques for automated detection of small metal particles in plastic food container packages. The sample packages consist of sealed cardboard boxes containing stacks of food containers: plastic cups for food, and Styrofoam cups for noodles. The primary goal of the investigation was to automatically identify small metal particles down to 0.5 mm diameter in size or less, randomly located within the containers. The multiple container stacks in each box make it virtually impossible to reliably detect the particles with single-energy X-ray techniques either visually or with image processing. The stacks get overlaid in the X-ray image and create many indications almost identical in contrast and size to real metal particles. Dual-energy X-ray techniques were investigated and found to result in a clear separation of the metal particles from the food container stack-ups. Automated image analysis of the resulting images provides reliable detection of the small metal particles.