Garuda - Garba Rujukan Digital

PENGENALAN WAJAH SECARA REAL TIME MENGGUNAKAN METODE CAMSHIFT, LAPALCIAN of GAUSSIAN DAN DISCRETE COSINE TRANSFORM TWO DIMENSIONAL (LoGDCT2D) Sultoni, Sultoni; Dachlan, Hary Soekotjo; Mudjirahardjo, Panca; Rahmadwati, Rahmadwati
Network Engineering Research Operation [NERO] Vol 2, No 3 (2016): NERO
Publisher : Universitas Trunojoyo Madura

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

Penelitian pengenalan wajah sudah mulai berkembang beberapa dekade terakhir dan sudah banyak diaplikasikan terutama untuk sistem keamanan. Banyak metode yang bisa digunakan dalam pengenalan wajah, salah satunya adalah pengenalan wajah secara real time dengan menggunakan metode Camshift, Laplacian of GaussianÂ dan Discrete Cosine Transforom Two Dimensional (LoGDCT2D). Berdasarkan hasil uji coba terhadap 10 orang dimana tiap-tiap orang diambil 10 citra wajah sebagai data pelatihan dengan berbagai posisi dan pose, kemudian dibandingkan dengan 100 frameÂ videoÂ orang yang bersangkutan, menghasilkan akurasi yang cukup baik apabila dibandingkan dengan metode DCT2D dan Gabor Wavlet. Berdasarkan penelitian terdahulu akurasi dari DCT2D adalah sekitar 85,40% dan Gabor Wavlet memiliki akurasi 79,31% sedangkan LoGDCT2D memiliki tingkat akurasi sebesar 93,91%. Dilihat dari waktu komputasi juga masih cukup baik, dimana rata â rata waktu komputasi dari LoGDCT2D adalah 2,14 detik ini mengindikasikan bahwa dengan adanya penambahan filtering menggunakan LoG dapat meingkatkan akurasi pengenalan dan dengan waktu komputasi yang tetap stabil.Kata kunci: Pengenalan Wajah, Camshift, LoG, DCT2D, LoGDCT2D.

PENGARUH VARIASI DYE CAROTENE DAN PHYCOCYANIN BERBAHAN DASAR DAUCUS CAROTA DAN SPIRULINA TERHADAP KARATERISTIK SENSOR OPTIK Rachmawati, Luthfiyah; Mudjirahardjo, Panca; Maulana, Eka
Jurnal Mahasiswa TEUB Vol 6, No 2 (2018)
Publisher : Jurnal Mahasiswa TEUB

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

Penelitian ini merancang sensor optik dengan stimulusÂ berupa cahaya lampu merkuri 160 Watt danÂ memberikanÂ keluaran berupa tegangan dan arus, perancangan sensor optikÂ menggunakan bahan dasar organik yaitu menggunakan dye carotene dan phycocyanin. Fabrikasi sensor optik pada penelitian ini menggunakan metode deposisi spin coating. Berdasarkan pengujian dye carotene memiliki puncak absorbansi cahaya pada 474 nm sebesar 2,882 a.u. Dye phycocyanin memiliki puncak absorbansi cahaya 620,50 nm sebesar 2,787 a.u.Â Pada pengujian tegangan dan arus sensor,Â makin tinggi iluminasi cahaya makin tinggi pula arus dan tengangannya. Sensor phycocyanin memiliki performansi yang paling baik diantara sensor lainnya dengan sensitivitas tegangan sebesar 5,0044mV/lux, sensitivitas arus 5,7524 ÂµA/lux, respon waktu 1,088 s, tingkat linieritas tegangan 77,54% dan tingkat linieritas arus 96,56%. Kata kunci : Sensor Optik, Spin Coating, Dye Carotene, Dye Phycocyanin. Â ABSTRACT This research is about designing optical sensor with light of mercury lamp 160 watt as input and the output are voltage and current, the design of optical sensor is using organic material such as dye carotene and phycocyanin. The fabrication of optical sensor in this research using spin coating as deposition method. Based on the experiment dye carotene had light absorbance peak at 474 nm in the amount of 2,882 a.u. On the voltage experiment and current sensor, the bigger light luminance resulting bigger current and voltage. Phycocianin sensor had the best performance among other sensor which sensitivity of voltage reach 5,0044mV/lux, sensitivity of current reach 5,7524 ÂµA/lux, response timeÂ 1,088 s, linearity voltage level 77,54% and liniarity current level at 96,56% Keywords: Optical Sensor, Spin Coating, Dye Carotene, Dye Phycocyanin.

IMPLEMENTASI FILTER GRAY LEVEL CO-OCCURANCE MATRIKS TERHADAP SISTEM KLASIFIKASI KANKER PAYUDARA DENGAN METODE CONVOLUTIONAL NEURAL NETWORK Rohman, Muhammad Ariefur; Mudjirahardjo, Panca; Muslim, M. Aziz
Transmisi: Jurnal Ilmiah Teknik Elektro Vol 23, No 4 Oktober (2021): TRANSMISI
Publisher : Departemen Teknik Elektro, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/transmisi.23.4.%p

Â Kanker payudara merupakan salah satu penyakit dengan proyeksi kematian terbesar selama 10 tahun terakhir dengan indeks kematian mencapai rata-rata 5 juta per-tahun, dan diprediksi akan terus naik hingga 60% di seluruh dunia. Umumnya, banyak metode yang digunakan untuk mendeteksi penyakit ini, salah satunya dengan mengamati jaringan histopatology. Banyak dari para ilmuwan, yang menggunakan jaringan histopatology untuk menganalisa, mengamati dan membuat sistem klasifikasi kanker payudara dengan berbagai metode, seperti: convolutional neural network, deep learning, support vector machine. Penggunaan metode convolutional neural network terbukti paling unggul pada sistem klasifikasi kanker payudara, namun akurasi rata-rata yang dihasilkan relatif cukup rendah. Selain itu, penggunaan metode convolutional neural network, membutuhkan waktu komputasi yang relatif lama untuk mengklasifikasikan 7909 dataset ukuran 4 GB. Berdasarkan alasan tersebut, desain sebuah sistem klasifikasi dengan mengimplementasikan Gray Level Co-occurance Matrix pada saat prapengolahan data CNN di butuhkan. Hasil penelitian menunjukkan bahwa, penggunaanÂ metodeÂ CNNÂ menghasilkanÂ waktuÂ komputasiÂ lebihÂ lama,Â yaitu:Â 3300Â detikÂ dibandingkanÂ dengan kombinasi metode GLCM Entropy dan CNN 2040 detik. Sedangkan rata-rata akurasi latih dan uji yang dihasilkan oleh metode kombinasi GLCM entropy dan CNN adalahÂ 92,26% dan 94,16%, lebih unggul dibandingkan dengan metode CNN, yaitu: 88,41% untuk data latih, dan 87,68% untuk data uji.

Comparison of Solar Radiation Intensity Forecasting Using ANFIS and Multiple Linear Regression Methods Hadi Suyono; Rini Nur Hasanah; R. A. Setyawan; Panca Mudjirahardjo; Anthony Wijoyo; Ismail Musirin
Bulletin of Electrical Engineering and Informatics Vol 7, No 2: June 2018
Publisher : Institute of Advanced Engineering and Science

Solar radiation forecasting is important in solar energy power plants (SEPPs) development. The electrical energy generated from the sunlight depends on the weather and climate conditions in the area where the SEPPs are installed. The condition of solar irradiation will indirectly affect the electrical grid system into which the SEPPs are injected, i.e. the amount and direction of the power flow, voltage, frequency, and also the dynamic state of the system. Therefore, the prediction of solar radiation condition is very crucial to identify its impact into the system. There are many methods in determining the prediction of solar radiation, either by mathematical approach or by heuristic approach such as artificial intelligent method. This paper analyzes the comparison of two methods, Adaptive Neuro Fuzzy Inference (ANFIS) method, which belongs into the heuristic methods, and Multiple Linear Regression (MLP) method, which uses a mathematical approach. The performance of both methods is measured using the root mean square error (RMSE) and the mean absolute error (MAE) values. The data of the Swiss Basel city from Meteoblue are used to test the performance of the two methods being compared. The data are divided into four cases, being classified as the training data and the data used as predictions. The solar radiation prediction using the ANFIS method indicates the results which are closer to the real measurement results, being compared to the the use MLP method. The average values of RMSE and MAE achieved are 123.27 W/m2 and 90.91 W/m2 using the ANFIS method, being compared to 138.70 W/m2 and 101.56 W/m2 respectively using the MLP method. The ANFIS method gives better prediction performance of 12.51% for RMSE and 11.71% for MAE with respect to the use of the MLP method.

Comparison of Solar Radiation Intensity Forecasting Using ANFIS and Multiple Linear Regression Methods Hadi Suyono; Rini Nur Hasanah; R. A. Setyawan; Panca Mudjirahardjo; Anthony Wijoyo; Ismail Musirin
Bulletin of Electrical Engineering and Informatics Vol 7, No 2: June 2018
Publisher : Institute of Advanced Engineering and Science

Solar radiation forecasting is important in solar energy power plants (SEPPs) development. The electrical energy generated from the sunlight depends on the weather and climate conditions in the area where the SEPPs are installed. The condition of solar irradiation will indirectly affect the electrical grid system into which the SEPPs are injected, i.e. the amount and direction of the power flow, voltage, frequency, and also the dynamic state of the system. Therefore, the prediction of solar radiation condition is very crucial to identify its impact into the system. There are many methods in determining the prediction of solar radiation, either by mathematical approach or by heuristic approach such as artificial intelligent method. This paper analyzes the comparison of two methods, Adaptive Neuro Fuzzy Inference (ANFIS) method, which belongs into the heuristic methods, and Multiple Linear Regression (MLP) method, which uses a mathematical approach. The performance of both methods is measured using the root mean square error (RMSE) and the mean absolute error (MAE) values. The data of the Swiss Basel city from Meteoblue are used to test the performance of the two methods being compared. The data are divided into four cases, being classified as the training data and the data used as predictions. The solar radiation prediction using the ANFIS method indicates the results which are closer to the real measurement results, being compared to the the use MLP method. The average values of RMSE and MAE achieved are 123.27 W/m2 and 90.91 W/m2 using the ANFIS method, being compared to 138.70 W/m2 and 101.56 W/m2 respectively using the MLP method. The ANFIS method gives better prediction performance of 12.51% for RMSE and 11.71% for MAE with respect to the use of the MLP method.

Optimasi Struktur Convolutional Neural Network LeNet5m dengan Pendekatan MorphNet Ridho Herasmara; Muhammad Aziz Muslim; Panca Mudjirahardjo
Jurnal EECCIS Vol 13, No 3 (2019)
Publisher : Fakultas Teknik, Universitas Brawijaya

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

Pendekatan perancangan neural network saat ini, menghasilkan rancangan yang tidak efisien. Rancangan yang tidak efisien ini menyebabkan penggunaan sumber daya yang lebih tinggi dibandingkan network yang lebih efisien. Hal ini juga merupakan permasalahan yang dialami network LeNet5, sebuah convolutional neural network untuk klasifikasi digit tulisan tangan yang dilatih dengan menggunakan dataset MNIST. Kami mengusulkan pendekatan MorphNet untuk optimasi kebutuhan flops-nya. Pendekatan MorphNet mengerdilkan network dengan menggunakan L1 regularization untuk menonaktifkan neuron pada tingkat aktivasinya. Neuron yang tidak aktif ini memiliki imbas yang kecil terhadap kinerja network, sehingga akan diusulkan untuk dihilangkan pada struktur yang baru. Network ini kemudian dapat dibesarkan untuk realokasi sumber daya. Sebagai hasilnya, didapatkan beberapa network baru yang lebih efisien dalam kebutuhan flops hingga 69%, dengan tetap mempertahankan tingkat akurasi pada rentang 98.5%. Kami menyimpulkan bahwa pendekatan MorphNet berhasil meningkatakan efisiensi dengan cara menghilangkan neuron yang berimbas kecil terhadap kinerja network.

Identifikasi Tanda Tangan dengan Ekstraksi Ciri GLCM dan LBP yuliana diah pristanti; Panca Mudjirahardjo; Achmad Basuki
Jurnal EECCIS Vol 13, No 1 (2019)
Publisher : Fakultas Teknik, Universitas Brawijaya

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

Signature identification with extraction features of GLCM (The Gray Level Co-occurrence Matrix) and LBP (The Local Binary Pattern) compare the results of both accuracy. By using signatures from 15 people, each of which has 10 signatures. For the training data, 7 signatures from each person were taken so that the training data amounted to 105 signatures. While the testing data was taken 3 signatures from each person so that the test data amounted to 45 signatures. From the results of image processing obtained the percentage using GLCM feature extraction is greater than the percentage using LBP feature extraction, namely GLCM reaches 86.67% and LBP 80.00%. But both remain at a high level of success. So it can be concluded that both GLCM and LBP feature extraction can be recommended to recognize signature textures. Index Terms—GLCM, LBP, Signature.

Comparison of Solar Radiation Intensity Forecasting Using ANFIS and Multiple Linear Regression Methods Hadi Suyono; Rini Nur Hasanah; R. A. Setyawan; Panca Mudjirahardjo; Anthony Wijoyo; Ismail Musirin
Bulletin of Electrical Engineering and Informatics Vol 7, No 2: June 2018
Publisher : Institute of Advanced Engineering and Science

Solar radiation forecasting is important in solar energy power plants (SEPPs) development. The electrical energy generated from the sunlight depends on the weather and climate conditions in the area where the SEPPs are installed. The condition of solar irradiation will indirectly affect the electrical grid system into which the SEPPs are injected, i.e. the amount and direction of the power flow, voltage, frequency, and also the dynamic state of the system. Therefore, the prediction of solar radiation condition is very crucial to identify its impact into the system. There are many methods in determining the prediction of solar radiation, either by mathematical approach or by heuristic approach such as artificial intelligent method. This paper analyzes the comparison of two methods, Adaptive Neuro Fuzzy Inference (ANFIS) method, which belongs into the heuristic methods, and Multiple Linear Regression (MLP) method, which uses a mathematical approach. The performance of both methods is measured using the root mean square error (RMSE) and the mean absolute error (MAE) values. The data of the Swiss Basel city from Meteoblue are used to test the performance of the two methods being compared. The data are divided into four cases, being classified as the training data and the data used as predictions. The solar radiation prediction using the ANFIS method indicates the results which are closer to the real measurement results, being compared to the the use MLP method. The average values of RMSE and MAE achieved are 123.27 W/m2 and 90.91 W/m2 using the ANFIS method, being compared to 138.70 W/m2 and 101.56 W/m2 respectively using the MLP method. The ANFIS method gives better prediction performance of 12.51% for RMSE and 11.71% for MAE with respect to the use of the MLP method.

Designing an Optimization of Orientation System toward Moving Object in 3-Dimensional Space Using Genetic Algorithm Feishal Reza; Panca Mudjirahardjo; Erni Yudaningtyas
JURNAL INFOTEL Vol 10 No 4 (2018): November 2018
Publisher : LPPM INSTITUT TEKNOLOGI TELKOM PURWOKERTO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20895/infotel.v10i4.408

This research offers a scheme of orientation system toward moving object in 3-dimensional space that using Stereo Vision Camera. The system benefits in giving an alternative solution in projecting practically without manual identification by conventional measuring device. The result of the projection in the system is in the form of coordinate position information (x, y, z), the length, the width, and the height of the object detected. The output displayed in the real-time digital image with 3-dimensional modeling. In the process of the object identification, there was a stage when an image was converted from colored image to binary image. But the conversion used the threshold method which was considered less efficient when an object moved. As consequence, the new adaptive method in solving the problem was needed. Genetic Algorithm was proposed as the optimization method because it was considered suitable with the emerging problems. In the optimization process, genetic algorithm was in a task of searching process and determining the threshold value as the process of creating binary image. The result shows an increased accuracy in the identification process after the system had been optimized by the Genetic Algorithm (GA).

Breast Cancer Detection using Residual Convolutional Neural Network and Weighted Loss Samuel Aji Sena; Panca Mudjirahardjo; Sholeh Hadi Pramono
JURNAL INFOTEL Vol 11 No 2 (2019): May 2019
Publisher : LPPM INSTITUT TEKNOLOGI TELKOM PURWOKERTO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20895/infotel.v11i2.430

This research presents a breast cancer detection system using deep learning method. Breast cancer detection in a large slide of biopsy image is a hard task because it needs manual observation by a pathologist to find the malignant region. The deep learning model used in this research is made up of multiple layers of the residual convolutional neural network, and instead of using another type of classifier, a multilayer neural network was used as the classifier and stacked together and trained using end-to-end training approach. The system is trained using invasive ductal carcinoma dataset from the Hospital of the University of Pennsylvania and The Cancer Institute of New Jersey. From this dataset, 80% and 20% were randomly sampled and used as training and testing data respectively. Training a neural network on an imbalanced dataset is quite challenging. Weighted loss function was used as the objective function to tackle this problem. We achieve 78.26% and 78.03% for Recall and F1-Score metrics, respectively which are an improvement compared to the previous approach.

Title

Found 10 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 10 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 10 Documents
Search