This study aims to classify the disease image on grape leaves using image processing. The segmentation uses the k-means clustering algorithm, the feature extraction process uses the VGG16 transfer learning technique, and the classification uses CNN. The dataset is from Kaggle of 4000 grape leaf images for four classes: leaves with black measles, leaf spot, healthy leaf, and blight. Google images of 100 pieces were also used as test data outside the dataset. The accuracy of the CNN model training is 99.50 %. The classification yields an accuracy of 97.25 % using the test data, while using test image data outside the dataset obtains an accuracy of 95 %. The designed image processing method can be applied to identify and classify disease images on grape leaves.

AbstrakMenyesuaikan tingkat kesulitan dengan perilaku pemain pada permainan menjadi penunjang agar pemain tidak berhenti bermain karena permainan yang dirasa terlalu sulit. Algoritma Genetika dapat digunakan sebagai metode untuk penyesuaianya. Metode ini berkerja dengan menyesuaikan tingkat kesulitan permainan sesuai dengan nilai fitness seorang pemain. Sayangnya, jika target nilai fitness mendekati ambang batas maksimal, algoritma ini membutuhkan proses iterasi yang banyak, bahkan memungkinkan terjadinya konvergensi prematur. Konvergensi prematur ini terjadi karena kurangnya keragaman populasi, sehingga nilai fitness minimal yang ditetapkan tidak memungkinkan untuk dicapai. Penelitian ini mencoba menginterupsi konvergensi prematur pada algoritma genetika agar tidak terjadi pengulangan iterasi yang tak berkesudahan dengan menetapkan dan menghitung batas keragamanya menggunakan Shanon-Wiener Diversity Index. Hasilnya, ketika ambang batas keragaman tercapai, nilai fitness saat keadaan itu diambil sebagai nilai paling optimum, sehingga tidak terjadi pengulangan iterasi yang tak berujung. Kata kunci-Konvergensi prematur, algoritma genetika, Shanon-Wiener Diversity Index AbstractAdjusting the level of difficulty with the behavior of players in the game to be supportive so that players do not stop playing because the game is considered too difficult. Genetic Algorithms can be used as a method for adjustments. This method works by adjusting the difficulty level of the game according to the value of a player's fitness. Unfortunately, if the target fitness value approaches the maximum threshold, this algorithm requires a lot of iteration processes, even allowing premature convergence to occur. This premature convergence occurs because of a lack of population diversity, so the minimum fitness value that is set is not possible to achieve. This study attempts to interrupt premature convergence of genetic algorithms to avoid endless iterations by setting and calculating diversity limits using the Shanon-Wiener Diversity Index. As a result, when the diversity threshold is reached, the fitness value when the state is taken as the most optimum value, so that there are no endless iterations.