Bartle, C.M.; West, J.G. 2009 Feasibility of DEXA prediction of dry matter and mass for horticultural products. Lower Hutt, N.Z.: GNS Science. GNS Science report 2009/47 18 p.
Abstract: Previously the DEXA system used in this research has been deployed commercially in meat processing plants with the objective of accurately determining the fat content of bulk and packaged meat, through a JV with ANZCO Foods and a partnership with Smiths Detection. This present research is aimed at demonstrating an ability to measure the dry matter distribution and in turn the net dry matter for individual horticultural products using DEXA. The DEXA images are obtained while scanning the products on a conveyor belt running at speeds representative of production grading situations. The products reported on here are primarily potatoes (because of a direct commercial interest), but also a butternut pumpkin and two rock melons. The grading and dry matter measurement capability is based on detecting change in the effective atomic number (EAN) with change in the elemental proportions within the product and there being effectively a binary mixture (e.g dry matter and water) present. Grading of fruit and vegetables on this basis is expected to be challenging. The commercial meat/fat grading already commercialised as a DEXA system is associated with 1.8 EAN units difference between fat (carbon rich Zeff=5.8) and fully-lean meat (oxygen rich Zeff.=7.6) but this range is large compared to what is expected for horticultural products. The dry matter in horticultural products is primarily starch (plus minerals) and the EAN difference between starch (and minerals) and water is unknown here but calculations give the difference as little as 0.2 EAN units, dependant on the mineral content in the product. In this work we show that the dry matter sensitivities of the DEXA technology for horticultural products is discernable allowing measurement of dry matter distributions, and net dry matter values. The EAN range is indeed much smaller than for fat/meat mixtures, and consistent differences are yet to be demonstrated for an assembly of product, except for potatoes where consistency is seen over a group of 15 potatoes. Further development of the DEXA imaging systems is needed to achieve the extra consistency needed across a full detector imaging array such as used across a conveyor. (auth)