@bsi.ac.id
Universitas Bina Sarana Informatika
Computer Science, Artificial Intelligence, Signal Processing, Computer Science Applications
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Sumanto, Adi Supriyatna, Irmawati Carolina, Ahmad Yani, Ruhul Amin, and Eka Dyah Setyaningsih
AIP Publishing
Yuni Sugiarti, Salma Riyanti Hanifah, E. Oos M. Anwas, Sumanto, Saipul Anwar, Anggraeni Dian Permatasari, and Evy Nurmiati
IEEE
Usability testing conducted on the PT Beli Jelantah Trafiguras website was carried out with the aim of identifying and analyzing the level of usability of the website, finding various new problems related to usability that are felt directly by website users, as well as providing recommendations for improvements in the form of analysis that can improve design and functionality on linked websites. The method used in this research is the Cognitive Walkthrough method with 10 (ten) respondents. The results of this study indicate that the related website still has several interface problems, such as unclear colors and font sizes, too many and ambiguous logos, incomplete registration instructions, and so on. The linked website also has a medium usability level with an effectiveness level of 67.5% and an efficiency level of 57.9%.
Sumanto, Bambang Wijonarko, Muhammad Qommarudin, Aji Sudibyo, Pudji Widodo, and Afit Muhammad Lukman
IEEE
Face detection has been one of the most explored problems in computer vision for several years. Using the WIDER FACES data set, this study investigates how the Viola-Jones method can be used to identify faces in 179 photos and how it performs compared to other face detection algorithms. In a previous study for face detection using Viola-jones, the highest accuracy results were obtained at 90.9% for facial images and 75.5% for non-face images. In this study, the Viola-Jones approach had a 100 percent success rate. This approach will be used in the MATLAB algorithm for face identification to get better results than currently available. Experiments using two classes had promising results.
Sumanto, Yuni Sugiarti, Adi Supriyatna, Irmawati Carolina, Ruhul Amin, and Ahmad Yani
IEEE
Apples are one of the most productive varieties of fruit in the world, with a high nutritional and medicinal value. However, numerous diseases affect apple production on a wide scale, resulting in significant economic losses. These diseases often go overlooked until just before, after, or after fruit has been processed. Many pathogens can be avoided with cultural traditions and (optional) fungicides, even if there are no cures for tainted fruit. However, accurate diagnosis is essential for determining the right management practices and preventing further losses. Apple scab, apple rot, and apple blotch are some of the most prevalent diseases that affect apples. The proposed approach will greatly aid in the automated identification and classification of apple diseases, according to our test results. We discovered that normal apples were easy to discern from diseased apples in our trial, and that the texture-based GLCM function produced more reliable results for apple disease classification, with a classification accuracy of more than 96.43 percent. This demonstrates that combining the GLCM extraction function with naive bayes classification will greatly improve accuracy.
Diah Puspitasari, Mochamad Wahyudi, Muhammad Rizaldi, Acmad Nurhadi, Kresna Ramanda, and Sumanto
IOP Publishing