Peaches decay quickly after harvest at ambient temperature due to their fast ripening and lack of anti-decomposition agents. Low temperature storage (at 2-7 °C) is an effective way to slow these decay processes and maintain crop quality. However, peaches are sensitive to low temperature, which can cause physiological disorders of fruit flesh, usually called chill damage or cold injury (CI), during cold storage, at market, or at home.
The symptoms of CI in peach fruit include internal browning and mealy or wooly flesh, characterized by a dry and mealy texture, lack of taste, aroma and extractable juice. CI limits the storage period and shelf life of fruit, thus reducing consumer acceptance and economic value.
Early detection and monitoring of CI in cold storage conditions is difficult, as the injured peach fruits often don't show damage as long as they are kept at a low temperature. The symptoms will develop and become evident immediately or over several days once peaches are transferred to a warm environment ( at 20°C or above); thus, the problem is commonly not noticed until the fruit reaches consumers. To remove defective fruits from the marketing chain earlier, a rapid, precise, reliable and non-destructive technique to detect the cold-injured peaches is required.
Chinese scientists observed indicators of CI in cold-stored peaches establishing a hyperspectral imaging system to detect cold injury, and developed an artificial neural network (ANN) model for which eight (487, 514, 629, 656, 774, 802, 920 and 948 nm) optimal wavelengths were selected in the visible and short-wave near-infrared spectral region (400–1000 nm).
The specific objectives were:
To analyze the changes of quality parameters (CI index, firmness, extractable juice, soluble solid content (SSC), titratable acidity (TA) and chlorophyll content) during cold storage, as well as the relationship between the quality parameters and spectral response; for each fruit, CI index was scored according to a four-grade scale, where 0 = none, 1 = slight, 2 = moderate, and 3 = severe.
To select optimal wavelengths and develop models for the purpose of discriminating normal from cold-damaged peaches, plus peach quality parameters, via the ANN model.
A total of 240 peaches (Prunus persica cv. 'Xiahui 5') without defects were hand-harvested at firm-mature stage from a commercial orchard in Nanjing (China). The lot was randomly divided into two groups, one group of 140 peaches (Group A) was stored at 0°C and 90% relative humidity (RH) for 6 weeks, the other group of 100 peaches (Group B) was stored at 5°C, 90% RH for 4 weeks. At one week intervals, 20 fruits were removed from storage and assessed for CI and quality, then held at 20°C (80% RH) for 3 days to simulate shelf conditions, whereupon their hyperspectral images were obtained.
Scientists found that peach quality factors changed with the development of cold injury during post-harvest storage at 0°C or 5°C. Based on the CI index, peaches stored at 5°C for 1 week and those at 0°C for 2 weeks remained normal (CI index = 0), while peaches in other groups became cold-injured (CI index > 0).
Between normal and chill-damaged peaches, significant differences in fruit quality parameters and the spectral response to correlating selected wavelengths were observed. Correlation coefficients between CI index, SSC, TA, chlorophyll content and the spectral response of selected wavelengths were -0.587 to -0.700, 0.393 to 0.552, 0.510 to 0.751, and 0.574 to 0.773, respectively, for cold-stored peaches.
With optimal representative wavelengths as inputs for the ANN model, the overall classification accuracy of chill damage was 95.8% for all cold-stored samples. The ANN prediction models for quality parameters performed well, with correlation coefficients from 0.6979 to 0.9026.
Scientists conclude that this research demonstrates the feasibility of hyperspectral reflectance imaging technique for detecting cold injury.
