CD Skripsi
Implementasi Metode Convolutional Neural Network Untuk Klasifikasi Tingkat Kematangan Mangga Menggunakan R
Mango must ripe to be more tasty. This is because ripe mango more sweeter than unripe mango. Sometimes, farmer checks for ripeness so that it can be enjoyed by mangoes lovers. This checking process called classification. This research purpose to do ripe and unripe mango identification and formed CNN classification model that was used for new mango image data. This research started from taking image used DSLR camera and image data size will be changed to 32 x 32 pixels with the help of EBImage package. Then the CNN model was formed used R language programming by using Keras and Tensorflow package. The result of the research was CNN method could be used to classify the mango ripness by 108 mango training data image and 24 mango testing data image sample and showed accuracy level of 97,1% for training data and 87,5% for testing data classified to do mango ripeness identification.
Keywords : Convolutional Neural Network, Deep Learning, Keras, Image, Image
Preprocessing, Mango, R.
Tidak tersedia versi lain