A least-squares support vector machine (LS-SVM) based on fractal analysis and CIELab parameters for the detection of browning degree on mango (Mangifera indica L.)

This paper introduces a least-squares support vector machine (LS-SVM) classifier to detect the degree of browning on mango fruits as a function of fractal dimension (FD) and L∗a∗b∗ values. Our results showed that the best classification accuracy of browning degree was up to 100% using the LS-SVM classifier based on FD and L∗a∗b∗ (γ = 6.13, σ2 = 9.36). However, the correct classification rates of 85.19% and 88.89% were achieved for the LS-SVM models based on FD (γ = 1.13, σ2 = 5.52) and based on L∗a∗b∗ (γ = 6.68, σ2 = 2.44), respectively. Therefore, this study indicated the possibility of developing a potentially useful classification tool using the LS-SVM combined with FD and L∗a∗b∗ values for classifying the degree of browning on mango fruits during processing, storage and distribution.

► Fractal dimension (FD) as an indicator to detect browning degree on mango fruits. ► LS-SVM based on FD and CIELab is a potential tool for grading of mango fruits. ► LS-SVM can be used as a modeling tool in food classification.