Deep Learning Methods for Filter Extraction in Tomato fruits

Abstract

In effort to productively utilize the exponential growth of image analysis and learning capability of Neural Networks (NN), we present our work which is dedicated to developing and training a deep neural network to extract meaningful patterns from a set of labeled data i.e. making generalizations. We show that Deep Neural Networks (DNNs) can learn feature representations that can be successfully applied in a wide spectrum of application domains. We showed how DNNs are applied to classification problems, grading of fresh tomato fruits based on their physical qualities using supervised learning approach. We achieved a result of about 60% accuracy using our local dataset which is quiet reasonable than using other standardized dataset as in the case of other researchers. Additionally, we are very sure of getting better result by fine-tuning some of our parameters because out network learns to generalize as the number iterations increases and so also the accuracy of predictions.