Evaluating bruise susceptibility of ’Golden Delicious’ apples using hyperspectral scattering technique

Authors: ZHU, QIBING - Jiangnan University; GUAN, JIYU - Jiangnan University; HUANG, MIN - Jiangnan University; Lu, Renfu; MENDOZA, FERNANDO - Michigan State University

Submitted to: Postharvest Biology and Technology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/9/2015
Publication Date: 12/23/2015
Citation: Zhu, Q., Guan, J., Huang, M., Lu, R., Mendoza, F. 2015. Evaluating bruise susceptibility of ’Golden Delicious’ apples using hyperspectral scattering technique. Postharvest Biology and Technology. 114:86-89.

Interpretive Summary: Bruising is one of the main concerns in development of automated harvesting and postharvest handling systems for apple and other fruits. To mitigate bruising damage to apples, it is important to understand and quantify the bruise susceptibility of apples using an objective method. Conventional methods for evaluating bruise susceptibility, which is defined as the volume of bruised tissue caused by per unit impact energy (in joules), require impacting the fruit against a rigid surface. In this research, hyperspectral scattering, an optical technique that measures light scattering and absorption properties of food products, was used for nondestructive prediction of the bruise susceptibility of apples. Hyperspectral scattering images were collected for 300 ‘Golden Delicious’ apples over a three-week period of storage. Each apple was then subjected to impact test at one of three energy levels (i.e., 0.33, 0.66, and 1.11 joules). The bruise volume of each fruit was determined from the color image of each tested fruit, using a bruise volume estimation equation. Prediction models for bruise susceptibility were developed from the hyperspectral scattering data for each impact energy level as well as for the pooled data. The models gave relatively good predictions for the bruise susceptibility of apples with the correlation coefficients of 0.826-0.919. The research demonstrated that hyperspectral scattering can be used for nondestructive measurement of apple bruise susceptibility, which would provide useful information for harvest and postharvest handling of apples.

Technical Abstract: This research evaluated the potential of hyperspectral scattering technique for predicting the bruise susceptibility of apples. Spectral scattering images between 500 and 1,000 nm were acquired for 300 ‘Golden Delicious’ apples over a time period of three weeks after harvest, using a hyperspectral imaging system. Individual apples were then subjected to impact test by a pendulum ball at one of the three levels of impact energy, i.e., 1.11, 0.66, and 0.33 J. Relative mean reflectance was computed for the scattering profile of each wavelength over 10 mm scattering distance to predict bruise susceptibility. Bruise volume was estimated from the digital images of the bruised fruit, using a bruise volume estimation model. The bruise susceptibility of apples, determined as the ratio of bruise volume to impact energy, ranged between 353 and 880 mm**3/J with the mean value of 567 mm**3/J and the standard deviation of 122 mm**3/J, and it was dependent on the impact energy level - the lower the impact energy, the higher the bruise susceptibility. Partial least squares (PLS) models for bruise susceptibility were developed for each impact energy level as well as for the pooled data. Better predictions of bruise susceptibility were obtained from the PLS models for each impact energy level, with the correlation coefficient of prediction or R=0.848-0.919 and root mean square error of prediction or RMSEP= 32.4-50.7 mm**3/J. Lower prediction results were obtained for the pooled data (R=0.826 and RMSEP=69.7 mm**3/J). This research demonstrated that spectral scattering measurement could be used for nondestructive prediction of apple bruise susceptibility.