prediction of firmness in peach fruit by means of laser light backscattering imaging system

quality evaluation is one of the effective activities in trade, economy and the health of communities. for this purpose, non-destructive methods are increasingly used, as they are faster and more economical in comparison with destructive ones. this study investigated the feasibility of predicting firmness by laser light backscattering imaging system, as a new non-destructive method for one cultivar of peaches. thus, a laser imaging system was assembled for capturing backscattering images, which consisted of one laser diode at 650 nm. after taking images, essential information of intensity and texture based statistical features was obtained by image analysis techniques, to build two types of calibration models. non-linear regression and artificial neural network were developed in order to find the best prediction models. consequently, final models based on the non-linear regression, gave the highest correlation coefficient of r = 0.89 to predict firmness.