Article Per Year (5 Year)
p-Index From 2020 - 2025
0.408
P-Index
This Author published in this journals
All Journal IAES International Journal of Artificial Intelligence (IJ-AI) International Journal of Engineering, Science and Information Technology
Dede Herman Suryana
Universitas Gunadarma
Astronomy Biochemistry, Genetics & Molecular Biology Chemical Engineering, Chemistry & Bioengineering Chemistry Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Earth & Planetary Sciences Education Electrical & Electronics Engineering Energy Engineering Industrial & Manufacturing Engineering Library & Information Science Materials Science & Nanotechnology Mathematics Mechanical Engineering Physics Social Sciences Transportation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Classification of meat using the convolutional neural network Detty Purnamasari; Koko Bachrudin; Dede Herman Suryana; Robert Robert
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 12, No 4: December 2023
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v12.i4.pp1845-1853

Abstract

Every animal meat has different color and texture, for example, beef has a dark red color with a chewy texture, while pork has a pale red color and smooth fiber. A previous study has classified types of meat using gray level co-ocurrence matrix (GLCM), hue saturation value (HSV), and color intensity. In this research, we created meat classification between beef, pork, and horse meat using a convolutional neural network (CNN) develop in jupyter notebook, using the MobileNetV2 model, and 315 meat images as a dataset divided into 3 groups, 70% image for the training dataset, 20% image for the testing dataset, and 10% image for validation dataset. Before dividing the image into 3 groups, the image is resized to 224×224, and convert the color to grayscale. The model is trained with a training dataset, the epoch of 50, and Adam optimizer, the results show an accuracy of 93.15%.
Applying Artificial Intelligence to Classify the Maturity Level of Coffee Beans During Roasting Dede Herman Suryana; Wahyu Kusuma Raharja
International Journal of Engineering, Science and Information Technology Vol 3, No 2 (2023)
Publisher : Department of Information Technology, Universitas Malikussaleh, Aceh Utara, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v3i2.461

Abstract

Coffee is a highly popular beverage worldwide. The quality of coffee is often judged based on its aroma and taste. Good coffee quality is influenced by various parameters during the coffee bean roasting process. Roasting is a crucial step where green coffee beans are heated at high temperatures, undergoing chemical reactions such as hydrolysis, polymerization, and pyrolysis. The color changes during the roasting process are caused by melanoidin, which results from Maillard and caramelization reactions, also impacting the flavor profile. Therefore, it is essential to accurately classify the level of coffee bean maturity. In the development of supercomputer technology, particularly with high-speed GPU microprocessors and large memory capacities, artificial intelligence algorithms have been widely implemented in various applications. Research on smart machines has been conducted to create systems resembling human intelligence. One of its applications is in recognizing the maturity level of coffee beans during roasting. In this study, image segmentation using ROI (Region Of Interest) and RGB color features are utilized to identify the characteristics of each coffee bean image. Additionally, CNN (Convolutional Neural Network) is employed for the classification stage, and this model is implemented into an Android smartphone device to detect the type of coffee bean being roasted. After the training process with 100 epochs, the model achieved a loss of 0.12 and a training accuracy of 94.79%. The model is capable of classifying images from the test data with an average accuracy of 85.83% and a loss value of 0.35.
Co-Authors Detty Purnamasari Koko Bachrudin Robert Robert Wahyu Kusuma Raharja