Banana ripening is accompanied by measurable changes in the optical properties of the peel. Conventional VIS/NIR spectroscopy can quantify these changes accurately, but requires specialized and relatively expensive equipment. Recent developments in computational photography suggest that spectral information can potentially be reconstructed from ordinary RGB images acquired with a smartphone.
Building on previous work, in which a prototype for smartphone-based spectral analysis of banana ripeness is being developed, this thesis will investigate the accuracy, robustness and practical applicability of this approach. The student will experimentally compare smartphone-based measurements with reference optical spectroscopy for bananas at different stages of ripeness. Particular attention will be paid to reproducibility, illumination conditions, camera settings, banana-to-banana variability and the ability to extract relevant spectral indicators of ripening.
The ultimate objective is to assess whether a low-cost camera-based system could provide a practical tool for non-invasive monitoring of banana ripeness during storage and refrigerated maritime transport, and to identify the improvements required for a reliable operational prototype.