Claim Missing Document
Check
Articles

Found 5 Documents
Search

Face Detection in Complex Background using Scale Invariant Feature Transform and Haar Cascade Classifier Methods Damarsiwi, Dyah Kartika; Pambudi, Elindra Ambar; Fitriani, Maulida Ayu; Wibowo, Feri
Sinkron : jurnal dan penelitian teknik informatika Vol. 8 No. 2 (2024): Article Research Volume 8 Issue 2, April 2024
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/sinkron.v8i2.13556

Abstract

Face detection is a process by a computer system that can find and identify human faces in digital images or videos. One of the main challenges faced in the face detection process is the complex background. Complex backgrounds, such as many color combinations in the image, can interfere with the detection process. To overcome this challenge, this research uses a combination of two methods: Scale Invariant Feature Transform (SIFT) and Haar Cascade Classifier. Scale Invariant Feature Transform (SIFT) is a method used in image processing to identify and describe unique features in an image. The SIFT method looks for keypoint descriptors in images that can be used as a reference in comparing different images. After the keypoint descriptor is found with SIFT, the Haar Cascade Classifier method is used to detect faces in the image. Haar Cascade Classifier is a practical algorithm for object detection in images. After facial features are extracted with these two methods, the results are compared with the K-Nearest Neighbor (KNN) approach. This research involves the introduction of 28 color images with complex backgrounds. The results of combining these two methods produce an accuracy of 81.75%. This shows that combining these two methods effectively overcomes complex background challenges in face detection.
Aspect Based Sentiment Analysis Menggunakan Indobert Model Terhadap Review Pengunjung Objek Wisata Baturraden Febrianto, Dany Candra; Fitriani, Maulida Ayu; Afrad, Mahazam; Khadija, Mutiara Auliya
Melek IT : Information Technology Journal Vol. 10 No. 2 (2024): Melek IT: Information Technology Journal
Publisher : Informatics Department-Universitas Wijaya Kusuma Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30742/melekitjournal.v10i2.358

Abstract

Tourism significantly contributes to regional economic growth and enhances public welfare. Baturraden tourist attraction, located in Banyumas Regency, Central Java, is one of the destinations whose main attraction is nature tourism. Data on visitors to Baturraden tourist attraction over the past few years shows a good trend. To ensure long-term sustainability and enhance service quality, understanding visitor perceptions and experiences is crucial. This study employs Aspect-Based Sentiment Analysis (ABSA) to analyze visitor reviews of Baturraden. Utilizing the IndoBERT model, a deep learning architecture based on Bidirectional Encoder Representations from Transformers (BERT) specifically tailored for the Indonesian language, the research focuses on four key aspects: Attraction, Accessibility, Amenities, and Ancillary Services. Next stage, a pre-processing process is carried out which includes Case Folding, Cleansing, Tokenizing, Normalization, Stemming and Stopword. Model evaluation is conducted using a confusion matrix, assessing accuracy (94.61%), precision (83.22%), recall (96%), and F1-score (88.11). These results demonstrate the model's can classif reviews into the required aspects.A primary challenge encountered in this research involves analyzing reviews exhibiting diverse linguistic styles, including variations in language and dialect, as well as addressing the issue of imbalanced data distribution across the different aspects.
Rainfall forecasting using triple exponential smoothing for rice cultivation in lamongan, jawa timur Widyantri, Shafrila; Hakim, Dimara Kusuma; Pambudi, Elindra Ambar; Fitriani, Maulida Ayu
Journal of Soft Computing Exploration Vol. 6 No. 1 (2025): March 2025
Publisher : SHM Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52465/joscex.v6i1.519

Abstract

Rice cultivation is a major agricultural activity that is heavily influenced by weather conditions. Extreme weather events, such as heavy rainfall, can cause farmers' productivity to decline. Rainfall forecasts are important for farmers to help them make the right decisions in managing their farming businesses. This research aims to predict rainfall in Lamongan Regency, East Java province, and provide valuable information to rice farmers to plan the optimal planting season. The method used in this study is Triple Exponential Smoothing (TES), an effective forecasting technique for processing time series data with seasonal patterns. Monthly rainfall data for the last five years formed the basis of the forecast, with data sourced from NASA's Power Data Access Viewer. The analysis results include a Mean Absolute Percentage Error (MAPE) value of 97.559% for rainfall. This rainfall forecast can assist farmers in increasing rice productivity and minimizing the risk of crop failure due to unpredictable weather conditions. With the rainfall weather forecast, farmers are expected to know the suitable months for rice cultivation so that productivity increases
Pengembangan Sistem Pakar untuk Skrining Awal Penderita Penyakit Tuberkulosis Menggunakan Forward Chaining Sidqiyah, Elis Dhia; Hindayati, Mustafidah; Fitriani, Maulida Ayu; Hamka, Muhammad
Jurnal Informatika: Jurnal Pengembangan IT Vol 10, No 2 (2025)
Publisher : Politeknik Harapan Bersama

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30591/jpit.v10i2.8621

Abstract

Tuberkulosis (TB) merupakan penyakit menular yang menjadi tantangan besar bagi kesehatan masyarakat di Indonesia, dengan banyak kasus yang belum terdeteksi. Berdasarkan data dari Sistem Informasi Tuberkulosis (SITB) per 2 Oktober 2023, sekitar 36% dari total kasus TB belum terlaporkan, yang berpotensi menjadi sumber penularan di masyarakat. Banyak masyarakat yang kurang memahami gejala TB, sehingga tidak menyadari pentingnya melakukan deteksi dini dan sering kali terlambat dalam mencari penanganan yang tepat. Kriteria yang diperlukan untuk skrining TB meliputi gejala seperti batuk berkepanjangan, demam, berkeringat pada malam hari, penurunan berat badan, sesak napas, dan pembesaran kelenjar getah bening. Penelitian ini bertujuan untuk mengembangkan sistem pakar berbasis web yang dapat digunakan untuk skrining awal TB. Metode yang diterapkan dalam penelitian adalah forward chaining, menggunakan aturan logika IF-THEN untuk menentukan hasil berdasarkan gejala yang dimasukkan oleh pengguna. Hasil pengujian kesesuaian aturan menunjukkan bahwa semua aturan yang diterapkan dalam sistem dapat menghasilkan kesimpulan yang tepat berdasarkan kombinasi gejala yang dilaporkan. Pengujian ini dilakukan dengan menggunakan metode berbasis kasus uji, di mana setiap kombinasi gejala diuji untuk memastikan bahwa sistem memberikan keluaran yang sesuai. Selain itu, sistem ini dilengkapi dengan antarmuka yang intuitif, sehingga masyarakat dapat dengan mudah melakukan skrining awal TB. Pengembangan sistem pakar ini diharapkan dapat memberikan kontribusi yang signifikan dalam pengendalian TB di Indonesia.
Perbandingan MobileNetV2, DenseNet121, InceptionV3, dan Xception pada Klasifikasi Citra Panel Surya Bersih dan Berdebu Nugroho, Aswin Mulyo; Mustafidah, Hindayati; Fitriani, Maulida Ayu; Supriyono, Supriyono
JURIKOM (Jurnal Riset Komputer) Vol. 12 No. 4 (2025): Agustus 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/jurikom.v12i4.8688

Abstract

The buildup of dust on solar panels can greatly diminish energy output, lower system efficiency, and raise operational expenses. A productive way to tackle this problem is to utilize image classification through Convolutional Neural Network (CNN) techniques. This study examines the classification capabilities of four CNN models, namely MobileNetV2, DenseNet121, InceptionV3, and Xception, using transfer learning. These models leverage pre-trained weights from large datasets such as ImageNet to accelerate convergence and improve generalization. The dataset of images utilized in this research is obtained from Kaggle and includes pictures of both clean and dusty solar panels. The dataset was divided into training, validation, and testing subsets using a stratified approach to ensure balanced class distribution across all subsets. During training, class weighting was used to address potential class imbalance. The models were developed using TensorFlow with multi-GPU support, optimized using the AdamW optimizer, and fine-tuned to enhance performance. Model evaluation was conducted using accuracy, precision, recall, and F1-score metrics. Among all the architectures evaluated, the Xception model achieved the best performance with an accuracy of 90.52%, outperforming MobileNetV2 with an accuracy of 87.92%, DenseNet121 with 89.78%, and InceptionV3 which achieved 87.73%. These results indicate that modern CNN-based models can effectively recognize relevant visual patterns to detect dust on solar panels.