Abstract
To reduce the risk of consumers contracting foodborne illness produce processors clean and sanitize food contact surfaces prior to production. Current methods to monitor surface hygiene, such as ATPase and culture counting assays, require sub-sampling at select locations and a wait time. A need exists for a method to detect contaminants on entire surfaces in real-time. A hand-held hyper-spectral imaging device that acquires fluorescence or reflectance images was developed to detect contaminants not readily visible to the naked eye. Device components include a 12-bit monochrome CCD camera, a 20 nm tunable band-pass filter, violet LED's to generate fluorescence excitation, a touch-screen monitor for real-time visualization and device control, and a laptop to operate software. Software allows the operator to change the device's wavelength, gain, and sampling rate and to conduct immediate, off-line, image analyses. Preliminary tests in a commercial facility have resulted in detection of residual processing debris and other potential contaminants on produce processing surfaces after cleaning and sanitizing. After looking at hyperspectral data it was determined that fluorescence images at optical wavebands of 475, 520, and 675 nm could be used to achieve best detection. Contaminants were typically found in difficult to clean locations such as damaged cutting boards, belts, and machine crevices. These results demonstrate the potential of multispectral, fluorescence imaging to enhance post-cleaning inspection, and, in practice, helped improve routine cleaning procedures. A similar, more cost-effective, device could be employed by companies, regulatory agencies and/or third party auditors to monitor or conduct inspections for surface hygiene monitoring in the food and medical industries.