Non-Destructive Estimation of Different Maturity Stages of Fuji Apples Using a Combination of Spectral Data and Majority Voting Method Based on Different Hybrids of Artificial Neural Network

Pourdarbani, Razieh and Sabzi, Sajad and Kalantari, Davood Non-Destructive Estimation of Different Maturity Stages of Fuji Apples Using a Combination of Spectral Data and Majority Voting Method Based on Different Hybrids of Artificial Neural Network. University of Mohaghegh Ardabili.

Text 5گزارش نهایی طرح.pdf Download (995kB)

Abstract

Non-destructive determination of different stages of fruit ripening has several advantages such as selective robotic harvesting or doing fertilizing operation depending on the stage of ripening. In this regards, the purpose of the present study was to investigate the non-destructive estimation of different ripening stages of Fuji apple using color and spectral data. Majority voting method based on five hybrid artificial neural network classifiers namely genetic algorithm, particle swarm optimization algorithm, firefly algorithm, support vector machine, and k-nearest algorithm. The inputs of classifiers consisted of color data including the second channel of L*a*b* color space and the hue angle and the spectral data including wavelength ranging from 465 to 485 nm, 675 to 700 nm and 870 to 890 nm. To evaluate the reliability of the majority voting method, classification procedure was performed in 1000 iterations. The results stated that in the best state of training and using different types of inputs, the correct classification rate of the majority voting method was 100%. The correct classification rate of the majority voting method using different input data including color data, spectral data of 465 to 485 nm, 675 to 700 nm and 870 to 890 nm were 95.12%, 99.37%, 97.56% and 97.80% respectively.

Actions (login required)

View Item