@article{Hongboonmee_Jantawong_2020, title={Apply of Deep Learning Techniques to Measure the Sweetness Level of Watermelon via Smartphone}, volume={10}, url={https://ph02.tci-thaijo.org/index.php/JIST/article/view/240500}, DOI={10.14456/jist.2020.7}, abstractNote={<p>This research proposed the development of automated system for analyzing sweetness and watermelon varieties from photos using deep learning techniques for use on smartphones. To help the public who would like to know the names of varieties and sweetness of watermelons. The main components of the system include: (1) Modeling, classifying varieties and sweetness levels of watermelons with deep learning neural network through the TensorFlow library, using the InceptionV3 and MobileNet algorithms to compare image classification. In which the trainers are able to classify 4 types of images, each type of 100 images, and training our model for 500 rounds. The result shows that the model from the InceptionV3 algorithm has the same accuracy as the model from the MobileNet algorithm. The accuracy is 97.20%. Therefore considering the model size obtained from learning, it found that MobileNet model size is smaller than InceptionV3, so choose the model from MobileNet to develop the system further. (2) Using the model from MobileNet algorithm to develop application on smartphones, which developed by Android Studio program. Results of the user satisfaction test, found that the average satisfaction is 4.34, standard deviation 0.62, it at good level. In conclusion, this application is effective and can use in real life.</p>}, number={1}, journal={JOURNAL OF INFORMATION SCIENCE AND TECHNOLOGY}, author={Hongboonmee, Nattavadee and Jantawong, Nutthapong}, year={2020}, month={Jun.}, pages={59–69} }