Historically, food inspection has required trained technicians to perform visual screening operations. A drawback of manual visual inspection is that latent defects that are not visible to the human eye cannot be recognized. Newer approaches, such as automated inspection systems based on imaging technology, seek to more efficiently differentiate between good and defective product. These methods typically rely on color and spatial features to make proper classifications. Advanced CCD imaging solutions from QImaging® have been demonstrated to produce reliable inspection results via high-resolution spectral imaging.
Detection and Classification of Defects and Diseases on Raw French Fries
In 2005, Jacco Noordam, Willie van den Broek, and Lutgarde Buydens published a study that examined the utility of spectral imaging technology for food inspection. One of the goals of the study was to investigate whether spectral imaging improves discrimination between potato flesh, potato peel, and potato defects on French fries.
The researchers compared spectral imaging and red/green/blue (RGB) color imaging technologies. The spectral imaging system was based on a slit-scanning approach in which a line of spatial information is dispersed into spectral components across a square detector (a Peltier-cooled monochrome CCD camera from QImaging was used). In the approach used, a single exposure contains all the spectral information from a single line across the specimen; a series of images is collected in order to generate a complete data set with “x and y” spatial and spectral information at each pixel.