Thumbnail Image

NIR technologies for high speed fruit grading

Today’s consumers seek information about food products so they can purchase with confidence, even if it translates to higher priced items. It is the aim of this thesis to improve the instruments available for the prediction of food quality, in particular, to advance the technology used for the rapid, low cost, non-destructive analysis of the internal quality of fruit. This thesis contains both experimental and theoretical work covering two main areas of research: the first deals with issues related to the practical grading of fruit, the second with the NIR light penetration and distribution within a fruit. A ‘rotary grader’ has been built to enable the rapid, non-destructive assessment of fruit. Up to twenty fruit can be placed on a rotating disk, which can be spun at a controlled speed or moved to fixed positions for stationary studies. A variety of optical arrangements can be mounted to enable the acquisition of spectra from individual fruit in different ‘modes’ of operation. A comparison between the different optical arrangements is presented. It was found that transmission mode (where the light passes right through the fruit) performs better than reflectance mode (where the detected light comes primarily from near the surface of the fruit). This result has been previously predicted and has now been demonstrated with the equipment used in this work. Different angles of transmission are compared as well as different fruit orientations for their influence on fruit quality predictions. It was found that 90° transmission (where the light is detected at right angles to the illumination of the fruit) might work best for randomly orientated kiwifruit; if they are hand placed then 180° or 150° may be more effective. A study of mandarins suggested that the acid content cannot be assessed directly, but it can be inferred from surface properties such as colour and chlorophyll content. This is contrary to several published reports and questions the ability of NIR spectroscopy to measure low concentrations of acid in fruit. A third experiment showed that soluble solids content (SSC) and dry matter (DM) can be assessed from moving apples with a prediction error of 0.54 ± 0.07 °Brix (R²ₚ = 0.65) and 0.66 ± 0.08 % (R²ₚ = 0.80) respectively. A unique perspective is presented showing where the useful information is attained from the continuously moving fruit. Transmission and reflectance spectra from apples moving at speeds of up to 3 fruit/s have been recorded, and the optimal spectra from the sequence were combined for improved predictions. To investigate the NIR light distribution and penetration in fruit a fibre optic probe was built to take direct measurements of the light levels inside fruit. This enabled, for the first time, the degree of light penetration to be assessed without altering the optical boundaries of the fruit (for example cutting away sections of the fruit). Corresponding simulations using Monte Carlo photon tracing, match the experimental measurements. It was found that fruit such as mandarin exhibit strong internal reflection from the skin, a result that has not previously been reported. The influence of the core and other features in the fruit on the light penetration and distribution is also studied showing perturbations in the internal light levels mapping out these features.
Type of thesis
Fraser, D. G. (2001). NIR technologies for high speed fruit grading (Thesis, Doctor of Philosophy (PhD)). The University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/14195
The University of Waikato
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.