Full text loading...
Abstract
The island of Malta has a rich heritage, which is evident in the rustic appeal of this island. Ġbejna forms an integral part of the Maltese food heritage. This artisan product is made from sheep or goat milk curds and aged for several months to develop its distinctive taste. During the ageing process, the cheese can become spoiled by fungi and unsafe for human consumption. This is a significant public health risk and a financial liability for producers. Conventional microbiology techniques do not detect these slow-growing pigment-less fungi early, allowing occasional distribution of contaminated products. We propose the use of hyperspectral imaging to detect these fungi during the early stages of cheese production. In contrast with a typical digital camera, which compiles the light signal into three broad wavelength bands; red, green and blue, a hyperspectral camera records numerous narrow and contiguous wavelength bands reflected from an object. This produces a series of images, each corresponding to the reflected electromagnetic energy in the respective narrow band of wavelengths. This image series may, in turn, be used for early fungal detection and identification. To test this hypothesis, a model cheeselet was produced to conduct compatibility and stability studies, through measurements of colony forming units, water activity, moisture levels, pH, protein and sugar content. The ġbejna model was then challenged with fungal strains isolated from commercial ġbejna and imaged using a hyperspectral camera. An algorithm is under development to differentiate contaminated samples from uncontaminated samples using image analysis and multivariate statistics.
- Published Online: