Fruit Recognition and Maturity Monitoring System

Abstract: In our nation, fruit recognition and its maturity monitoring is a difficult task due to the mass production of fruit products. In order to determine and evaluate the quality of the fruit accurately. The project presents, In real time with camera developing a sorting machine for classification of multiple fruits and checking the time they survive and evaluate the rank of the fruits based on its quality. Firstly, the algorithm bring out the RGB image values and the background of image was disengaged by the split-and-merge algorithm. Secondly, the extracted multiple features are namely its color, statistical, textural and geometrical feature. Geometrical features are used in the evaluation of quality of the fruits. Additionally four different classifiers k-nearest neighbour(k-NN), support vector machine(SVM), sparse representative classifier(SRC) and artificial neural network(ANN) are used to classify the fruits. The SVM classifier has seen to be more effective in quality evaluation. Using k-fold cross-validation techniques validates the system performance by considering different values of k. The classification is among Rank1, Rank2 and defected one. The system achieved maximum accuracy for fruit detection and classification.

Keywords: Classification, Defect detection, Geometrical feature, Textural feature, Statistical feature