Claim Missing Document
Check
Articles

Found 19 Documents
Search

MEMBUAT PIXEL ART MENGGUNAKAN LEARNING VECTOR QUANTIZATION AS, Putra Wisnu; Sumpeno, Surya; Sumpeno, Surya; Purnomo, Mauridhi Hery; Purnomo, Mauridhi Hery
Seminar Nasional Aplikasi Teknologi Informasi (SNATI) 2012
Publisher : Jurusan Teknik Informatika, Fakultas Teknologi Industri, Universitas Islam Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Pixel art didesain untuk banyak kepentingan dalam merepresentasikan wujud real life look seni visual artefak kontemporer. Pixel art adalah kratifitas seni membangun gambar dari unit kecil berwarna yang disebut dengan piksel, dari sebuah citra digital. Melukis pixel art pada bidang tulis digital dengan komposisi lukisan yang kompleks memiliki beberapa tantangan yang serius. Permasalahannya adalah penggunaan warna dalam pixel painting mengandung prasyarat bahwa pixel diisi degan warna yang penuh, mampu menggambarkan warna transisi serta dapat menunjukan corak warna terang dan gelap. Dari pemahaman ini, dipandang penting untuk membuat suatu metode melukis pixel art alternatif. Berperan sebagai kanvas digital, bidang citra diterjemahkan kedalam painting surface yang berkorespondensi terhadap nilai panjang dan lebar citra sedangkan kuantitas satuan piksel yang memaknai ruang data spasial dan warna adalah sebagai kuas maya. Goresan tinta diproduksi melalui serangkaian olah matematis dengan pendekatan vector quantization untuk diajarkan agar mampu merepresentasikan wujud citra alami yang realisitis. Penelitian ini menunjukan keberhasilan pengorganisasian parameter kedekatan jarak antara piksel dan kelas utamanya pada nilai 0.5 satuan piksel sebagai ruang optimum sehingga voronoi area dapat tercipta dengan baik. Sebagai bahan uji kulifikasi disediakan sejumlah citra pixel art yang berhasil ditransformasikan dengan baik berdasarkan hasil evaluasi area yang tersegmentasi lebih artistis ketika codebook yang disebarkan adalah proporsional dengan jumlah objek pixel art. Sehingga ikhtisar untuk mendapatkan corak pixel art dengan vector quantization adalah dengan memastikan jumlah codebook yang disebar berdasarkan referensi pola informasi spasial dan warna objek citra.
MEMBUAT PIXEL ART MENGGUNAKAN LEARNING VECTOR QUANTIZATION AS, Putra Wisnu; Sumpeno, Surya; Sumpeno, Surya; Purnomo, Mauridhi Hery; Purnomo, Mauridhi Hery
Seminar Nasional Aplikasi Teknologi Informasi (SNATI) 2012
Publisher : Jurusan Teknik Informatika, Fakultas Teknologi Industri, Universitas Islam Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Pixel art didesain untuk banyak kepentingan dalam merepresentasikan wujud real life look seni visual artefak kontemporer. Pixel art adalah kratifitas seni membangun gambar dari unit kecil berwarna yang disebut dengan piksel, dari sebuah citra digital. Melukis pixel art pada bidang tulis digital dengan komposisi lukisan yang kompleks memiliki beberapa tantangan yang serius. Permasalahannya adalah penggunaan warna dalam pixel painting mengandung prasyarat bahwa pixel diisi degan warna yang penuh, mampu menggambarkan warna transisi serta dapat menunjukan corak warna terang dan gelap. Dari pemahaman ini, dipandang penting untuk membuat suatu metode melukis pixel art alternatif. Berperan sebagai kanvas digital, bidang citra diterjemahkan kedalam painting surface yang berkorespondensi terhadap nilai panjang dan lebar citra sedangkan kuantitas satuan piksel yang memaknai ruang data spasial dan warna adalah sebagai kuas maya. Goresan tinta diproduksi melalui serangkaian olah matematis dengan pendekatan vector quantization untuk diajarkan agar mampu merepresentasikan wujud citra alami yang realisitis. Penelitian ini menunjukan keberhasilan pengorganisasian parameter kedekatan jarak antara piksel dan kelas utamanya pada nilai 0.5 satuan piksel sebagai ruang optimum sehingga voronoi area dapat tercipta dengan baik. Sebagai bahan uji kulifikasi disediakan sejumlah citra pixel art yang berhasil ditransformasikan dengan baik berdasarkan hasil evaluasi area yang tersegmentasi lebih artistis ketika codebook yang disebarkan adalah proporsional dengan jumlah objek pixel art. Sehingga ikhtisar untuk mendapatkan corak pixel art dengan vector quantization adalah dengan memastikan jumlah codebook yang disebar berdasarkan referensi pola informasi spasial dan warna objek citra.
Tracking System for Indoor TV Antenna Based on CVBS Signal Processing Miawarni, Herti; Hidayat, M. Mahaputra; Sumpeno, Surya; Setijadi, Eko
Jurnal Elektronika dan Telekomunikasi Vol 17, No 2 (2017)
Publisher : Indonesian Institute of Sciences

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14203/jet.v17.48-55

Abstract

Analog terrestrial TV is still a popular choice for urban societies although the migration to digital system has already begun. Video quality of analog TV was heavily influenced by performance of the antenna. Most users prefer to use indoor antenna due to its simplicity. The disadvantage of this type of antenna is the users may need to change the antenna direction repeatedly when they change to different TV channel. In this research, we designed and developed tracking system that enable indoor TV antenna to adjust its direction automatically to get optimum video clarity. This system is built by several servo motors and telescopic antennas. Composite Video Baseband Signal (CVBS) processing is used to obtain reference information regarding video clarity level conditions on TV screen. The results show that CVBS signal processing has performance in describing video clarity level. System performance has been verified from trial results on some UHF channels. Minimum tracking time is reach 23.4 second and the maximum reach 24.6 second.
Deteksi Gestur Lengan Dinamis pada Lingkungan Virtual Tiga Dimensi Koleksi Warisan Budaya Sooai, Adri Gabriel; Rumaksari, Atyanta. N.; Khamid, Khamid; Fanani, Nurul Zainal; Sumpeno, Surya; Purnomo, Mauridhi Hery
Jurnal Nasional Teknik Elektro dan Teknologi Informasi (JNTETI) Vol 7, No 4 (2018)
Publisher : Jurusan Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1256.2 KB) | DOI: 10.22146/jnteti.v7i4.457

Abstract

Virtual reality technology can be used to support museum exhibitions. Implementation could be in various platforms. There are many implementation options, for example in smartphones, tablet, and desktop computers. Most objects of museum collections are very fragile. Minimizing the direct touch on a collection object is one of the benefits of this technology. This study aims to prepare gestures suitable for the exploration of virtual objects of cultural heritage collection. Five sets of gestures have been prepared, namely lifting, picking, holding, sweeping from both directions, left and right. Dynamic arm gestures are recorded using the forearm sensor. The recorded data contains coordinates of gestures in form of x, y, z, raw, pitch, and yaw. Gaussian mixture models are used in selecting features to produce good accuracy in the classification process. Two functions are used, namely probability density function and cumulative distribution function for the feature selection process. In this study, two experiments were used to train the gesture model. The accuracy of the two experiments is shown in the form of a confusion matrix. Each of the confusion matrices show excellent results of 99.8% for SVM and k-NN. Furthermore, modeling results are tested using new data. The testing shows 89.25% result for SVM classifier and 90.09% for k-NN. Four other dynamic arm gestures have a very satisfactory rate of 100% for the two classifiers. The five gestures can be used in the development of virtual reality applications.
ANIMASI GERAKAN EXAGGERATION PUKULAN TINJU BERBASIS PENDEKATAN KURVA BEZIER Primasetya, Aidil; Sumpeno, Surya; Hariadi, Moch.
SENTIA 2015 Vol 7, No 2 (2015)
Publisher : SENTIA 2015

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (664.384 KB)

Abstract

Pukulan dapat digunakan sebagai objek gerak yang ditangkap dengan motion capture (mocap). Gerakan tersebut banyak di temukan pada olahraga tinju. Data mocap berisi informasi titik-titik koordinat yang mewakili rangka manusia. Titik-titik koordinat tersebut ketika disusun berdasarkan perubahan waktu akan membentuk lintasan gerakan. Exaggeration merupakan salah satu prinsip animasi 2D yang melebih-lebihkan gerakan atau ekspresi wajah. Prinsip ini tidak dapat secara langsung diaplikasikan ke dalam animasi 3D berbasis mocap. Karena membutuhkan pengambilan data yang berulang-ulang. Pada penelitian ini, data mocap dimodifikasi dengan menggunakan rotasi matriks. Modifikasi ini menghasilkan data mocap baru. Animasi gerakan pada data tersebut terlihat  exaggeration  dan  tidak  alami.  Untuk  mengembalikan  kealamian  gerakan  tersebut  maka  digunakan metode  interpolasi  yaitu  kurva  linear  bezier  dan  kurva  kuadratik  bezier.  Hasil  percobaan  menunjukkan perbedaan lintasan gerakan pada animasi gerakan exaggeration pukulan tinju menggunakan matriks rotasi yang diinterpolasi dengan kurva linear bezier dan kurva quadratik bezier.
SISTEM DETEKSI KEJERNIHAN VIDEO PADA TELEVISI ANALOG BERBASIS PENGOLAHAN SINYAL CVBS DAN PENDEKATAN MOS VQS Miawarni, Herti; Hidayat, M Mahaputra; Sumpeno, Surya; Setijadi, Eko
Prosiding SNATIF 2017: Prosiding Seminar Nasional Teknologi dan informatika (BUKU 1)
Publisher : Prosiding SNATIF

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

AbstrakTracking antena otomatis untuk televisi analog membutuhkan parameter acuan yang dapat digunakan sebagai referensi. Salah satu parameter yang berpotensi untuk dapat digunakan adalah sinyal keluaran AV-Out pada perangkat televisi. Sinyal video analog CVBS pada AV-Out diubah menjadi deret pulsa agar dapat diolah oleh prosessor. Hasil pengolahan akan menghasilkan data yang dapat mendeskripsikan tingkat kejernihan video pada layar televisi. Selanjutnya, data tersebut akan digunakan oleh antena dalam menentukan arah terbaik guna mendapatkan kejernihan video maksimum. Sebagai upaya realisasi, maka dibangun sistem deteksi kejernihan video pada televisi berbasis pengolahan sinyal analog CVBS (Composite Video Baseband Signal) dan menggunakan pendekatan subyektif MOS-VQS (Mean Opinion Score - Video Quality Subjective). Beberapa hasil uji coba menunjukan bahwa, rangkaian CVBS to Pulse Converter yang telah didesain mampu mengkonversikan sinyal analog CVBS menjadi deret pulsa berlevel tegangan TTL (Transistor Transistor Logic). Secara keseluruhan, sistem dapat bekerja dengan baik saat tingkat kejernihan video pada kondisi jernih, kabur dan noisy. Akurasi dapat dicapai 100% saat kondisi jernih dan noisy, sementara akurasi mencapai 75% saat kondisi kabur. Kata kunci: Televisi  Analog, Tracking Antena , AV-Out, CVBS, MOS-VQS
Ekstraksi Ciri Produktivitas Dinamis untuk Prediksi Topik Pakar dengan Model Discrete Choice Purwitasari, Diana; Fatichah, Chastine; Sumpeno, Surya; Purnomo, Mauridhi Hery
Jurnal Nasional Teknik Elektro dan Teknologi Informasi (JNTETI) Vol 7, No 4 (2018)
Publisher : Jurusan Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1928.298 KB) | DOI: 10.22146/jnteti.v7i4.460

Abstract

Recommendation of active or productive experts is indispensable in supporting collaborations. Activities of publication and citation indicate expert productivity. An expert can be inferred to have an interest in a subject through productivity in that particular topic. Since an expert can change interests over time, the contribution of this paper is a Discrete Choice Model (DCM) based on topic productivities to predict the primary interests of the experts. DCM uses features extracted from bibliographic data of citation relation and title-abstract texts. Before extracting productivity features and dynamicity features to represent interest changes, title clustering with KMeans++ is used to identify research topics. There are six productivity features and five dynamicity values for each productivity feature to demonstrate the expert behavior. Therefore, a clustered topic as a research interest is represented as an expert choice with 30 extracted features in the proposed method. The experiments used multinomial logistic regression for DCM and a log-likelihood indicator for the fitted models of the features. The resulted DCM models showed that productive behavior of the experts by doing many publications and receiving many citations effected to the precision of topic prediction by 80%. Some features were better for predicting primary interests of the expert. It was demonstrated with a lower precision value of 60% by using features that represent the expert behavior of only doing publication or only getting citation.
Performance IEEE 802.14.5 and zigbee protocol on realtime monitoring augmented reality based wireless sensor network system Editya, Arda Surya; Sumpeno, Surya; Pratomo, Istas
International Journal of Advances in Intelligent Informatics Vol 3, No 2 (2017): July 2017
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26555/ijain.v3i2.99

Abstract

The internet of Thing (IoT)technology has much development in this era. It has various wireless media transmission systems such as ESP and XBEE. Some IoT device can monitor website or application. On the other hand, Augmented Reality (AR) is a technology that used more on the entertainment sector. Here, we try to use AR to monitor the xbee based IoT device. As a result, there is the different result between Zigbee Protocol and IEEE 802.14.5 real time monitoring system. The optimum estimation of realtime time tolerance of those monitoring systems is >1500 ms (IEEE 804.14.5) and > 50 ms (Zigbee protocol).
Menuju Pengenalan Ekspresi Mikro: Pendeteksian Komponen Wajah Menggunakan Discriminative Response Map Fitting Rosiani, Ulla Delfana; Choirina, Priska; Sumpeno, Surya; P., Mauridhy Hery
Jurnal Nasional Teknik Elektro dan Teknologi Informasi (JNTETI) Vol 7, No 2 (2018)
Publisher : Jurusan Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1573.841 KB) | DOI: 10.22146/jnteti.v7i2.424

Abstract

The observations made in the study of micro-expression are to recognize and track the very subtle movements of certain facial areas and in a short time. In this study, the observation of movement is held in some areas of the face component. The facial and facial components detection is the pre-process stage on micro-expression recognition system. The goal at this stage is to get face and face components accurately and quickly on every movement of the video sequence or image sequence. The face landmark point of the Discriminative Response Map Fitting (DRMF) method can be used to get face components area accurately and quickly. This can be done because the facial landmark points used in this model-based method do not change when objects are moved, rotated, or scaled. The results obtained by using this method are accurate with a 100% accuracy value compared to the Haar Cascade Classifier method with an average accuracy of 44%. In addition, the average time required in the formation of facial component boxes for each frame is 0.08 seconds, faster than the Haar Cascade Classifier method of 0.32 seconds. With the results obtained, then the detection of facial components can be obtained accurately and quickly. Furthermore, the boxes of face components obtained are expected to display the appropriate data to be processed correctly and accurately in the next stage, feature extraction and the classification of micro-expression motion stage.
Analisis Kinerja LSTM dan GRU sebagai Model Generatif untuk Tari Remo Zaman, Lukman; Sumpeno, Surya; Hariadi, Mochamad
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 8, No 2 (2019)
Publisher : Jurusan Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1418.245 KB) | DOI: 10.22146/jnteti.v8i2.503

Abstract

Creating dance animations can be done manually or using a motion capture system. An intelligent system that able to generate a variety of dance movements should be helpful for this task. The recurrent neural network such as Long Short-Term Memory (LSTM) or Gated Recurrent Unit (GRU) could be trained as a generative model. This model is able to memorize the training data set and reiterate its memory as the output with arbitrary length. This ability makes the model feasible for generating dance animation. Remo is a dance that comprises several repeating basic moves. A generative model with Remo moves as training data set should make the animation creating process for this dance simpler. Because the generative model for this kind of problem involves a probabilistic function in form of Mixture Density Models (MDN), the random effects of that function also affect the model performance. This paper uses LSTM and GRU as generative models for Remo dance moves and tests their performance. SGD, Adagrad, and Adam are also used as optimization algorithms and drop-out is used as the regulator to find out how these algorithms affect the training process. The experiment results show that LSTM outperforms GRU in term of the number of successful training. The trained models are able to create unlimited dance moves animation. The quality of the animations is assessed by using visual and dynamic time warping (DTW) method. The DTW method shows that on average, GRU results have 116% greater variance than LSTM’s.