A non-destructive estimation method for the soluble solids content of apples

As a non-destructive measurement method, the near-infrared spectroscopy (NIRS) has been widely employed in the agriculture field for estimating soluble solids content (SSC) of the fruit, which is a key parameter for consumers. A rapid and non-destructive estimation method is proposed in this work to measure the SSC of apples by using NIRS. A modified genetic algorithm (GA) is used for the systematic optimization of characteristic wavelengths. Multiple linear regression (MLR) is also adopted to find the coefficient vector. In order to evaluate the performance of the algorithm, numerical simulation and practical implement have been carried out. The results which based on low standard error of prediction (LSEP) and relatively high ratio to prediction (RPD) have shown the effectiveness and efficiency of the proposed method.