Claim Missing Document
Check
Articles

Found 9 Documents
Search

KLASIFIKASI OBJEK BERBASIS GAMBAR THERMAL MENGGUNAKAN DEEP LEARNING (PRE-TRAINED RESNET 50) Nugroho, Agung; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.201

Abstract

The development of technology in the field of transportation is currently something that is very enthusiastically welcomed by the Indonesian people in general. But along with the development of transportation technology that exists today is much different from the past, with a lot of sophistication and improving quality and safety that is more innovative. An Autonomous Car was created that can make it easier for drivers and maintain safety while driving. This system was built using the Neural Network control method, as well as Image Processing as the input signal in the form of images, and with the Flir camera as the vehicle data input. This of course has a very positive impact on human life today, of course humans will be more efficient in time, maintain safety on the trip, and can be more productive when driving. The method that is currently developing rapidly is automatic extraction using deep learning technology. The method used is Fully Convolutional Network (FCN) with Residual Neural Network (ResNet) architecture.The method currently used in the research is automatic extraction using deep learning technology to detect objects in the classification that has been made, with Residual Neural Network 50 (ResNet) architecture. In this study, the extraction accuracy for automatic vehicle function training reached 97.1% for ResNet 50 and 96.7% for ResNet 101 with a resolution of 640x512 pixels.
Peningkatan Kualitas Sistem Informasi Berbasis Website untuk Kendaraan Listrik Otonom Septiyanti, Riska Yucha; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.204

Abstract

Current technological developments have made the website a means of disseminating information that offers various conveniences and speeds in its presentation. Speed and convenience are positive values of the existence of the internet. The website is more easily accessible by people in various regions just by using the internet. One of the benefits of the website is that in terms of marketing the products of a business, the website can be a means of promoting, marketing and conveying information effectively and efficiently to the public. Another benefit of website technology is for information purposes of autonomous electric vehicles. This technology in the future can be integrated on the website for the purpose of information on its existence and condition. In fact, it is possible to create applications for the use of autonomous electric vehicles by making calls through the website, as is the case with applications that are currently popular, namely Gojek or Grab.In designing a website-based information system, a markup language consisting of HTML and CSS is used which is synchronized so that it can run simultaneously, besides that, the PHP programming language is also used to make the website look more dynamic and responsive. The method used in improving the quality of the website- based information system for autonomous electric vehicles is the static website method. On static websites the content provided is constant or does not change and on each page is made with HTML code.In this study, a website-based information system was developed for the purposes of a simple autonomous electric vehicle, which operates in the Nurtanio University Bandung campus. This website design consists of 8 pages. Page 1 contains a homepage where general information on research activities for the development of simple autonomous electric vehicles is found. Page 2 contains object detection which displays more detailed information about the object detection technology used in autonomous electric vehicles. Information about the presence of autonomous electric vehicles in the campus environment is then shown on page 3, namely Av Tracker. Page 4 contains the results of the research that has been done. Page 5 contains research opportunities. This website design is also equipped with a contributor page on page 6 which contains researchers involved in autonomous electric vehicle research. Page 7 contains biodata of alumni who have contributed. At the end, page 8 is for contact information for those who are interested in communicating further about matters related to this research.The final form of a website-based information system is to provide convenience to convey information about autonomous electric vehicles, and also to provide convenience for website management according to the required data and for further development with a tracking system.
Pemanfaatan Metoda RANSAC Untuk Validasi Rekontruksi Objek 2 Dimensi Dengan Menggunakan LiDAR Mulyana, Tri; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.205

Abstract

This research is a development of Random Sample Consensus (RANSAC) technology which is one of the facilities in autonomous type electric vehicles. Random Sample Consensus (RANSAC) is a two-dimensional technology that detects the presence of objects so that the vehicle can respond in the form of braking or maneuvering to avoid these objects. In the application of RANSAC, it is often constrained by anomaly data which affects the accuracy in the detection of actual objects. In this study, it aims to overcome anomaly data. In this study, it was assisted by MATLAB software which was used for analysis, comparison and programming whose results were entered into Excel as a dataset for the Reconstruction of 2-Dimensional Objects Using RANSAC Modifications The result data will be processed using the Linear Regression method or prediction based on previous data and using the Least Square method or the Least Squares method. The results of the study used samplesize and maxdistance which varied, from the first test getting 92.8% of the data declared good and the second test getting 58.33% of the data declared good. The results of this study show that the reconstruction of 2-Dimensional objects from LiDAR data can be validated using the RANSAC method, the Robust Fit line and the Least Square Fit from the image can be changed by setting SampleSize and MaxDistance, in the test results that have been carried out the tested data are declared successful, and the amount of data and the determination of the amount of data can affect the results in the study.
SEGMENTASI OBJEK BERBASIS GAMBAR TERMAL MENGGUNAKAN DEEP LEARNING (PRE-TRAINED RESNEXT 50) Fauzan, R. Aldam Dwi; Satyawan, Arief Suryadi; Siswanti, Sri Desy; Puspita, Heni
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.207

Abstract

The transportation sector at this time has experienced many technological developments which have been well receifed by the public, especially the people of Indonesia. Along with the development of transportation technology has undergone many developments, with sophistication and increased comfort and better security. So Autonomus Car technology was created that can help drivers to maintain safety while driving. Autonomus car was built using the Neural Network control method, and also Image Processing as signal processing with image input, and with a flip camera used for vehicle input data. Autonomous cars have many positive impacts on human life today, so humans can minimize time properly. Travel safety is maintained, and can be more productive when driving. The method that is currently developing rapidly is automatic extraction using deep learning. In this final project, automatic extraction method with deep learning technology used is Fully Convolutional Network (FCN) with Residual Neural Network Next (ResNext) architecture. In this study, the extraction accuracy for automatic vehicle function training reached 98% for ResNext 50 with a resolution of 640x540 pixels. Semantic segmentation will then test with 34030 image frames offline. In ResNext 50 architecture contains 20512 frames in good category, 7883 in adequate category and 5605 in poor category.
KLASIFIKASI BUMBU DAPUR DENGAN GAMBAR 360 DERAJAT (FISH EYE) MENGGUNAKAN TENSORFLOW Nurul P., Vethrea D. Gynandra; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.208

Abstract

Classification of objects is one of the studies that are currently being developed. One of the object classification techniques that are widely used is to implement deep learning methods. Convolutional Neural Network (CNN) is included in the type of Deep Learning because of the depth of the network. CNN is a convolution operation that combines several processing layers, using several elements that operate in parallel. Often people make the mistake of distinguishing one object from another, such as kitchen spices. Similar colors and shapes can make an error in the taste of food, therefore this object classification takes kitchen spices as an example to recognize and classify kitchen spices objects in making it easier for humans to recognize objects. The objects used are Ginger, Candlenut, Salam Leaves, Coriander and Lemongrass which are used as ingredients for classifying kitchen spice objects based on shape. Classification of objects that are classified using the CNN and Tensorflow methods. CNN construction will be built for object classification applications on 360° camera images (fish eye). An image in the form of a video of kitchen spices will be taken using a 360° camera and used as a model for classifying objects using Tensorflow and Jupyter Notebook for training data. The results of this detection system are expected to be able to work well in classifying objects in 360° image format that have significant distortion.
SEGMENTASI OBJEK BERBASIS KAMERA TERMAL MENGGUNAKAN DEEP LEARNING (PRE-TRAINED RESNET 34) Laksono, Muhammad Fauzan Anggi Fathul; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.210

Abstract

Technology is developing very rapidly in a pluralistic country, among these technologies are electric vehicles without human intervention, we are more familiar with this with autonomous electric vehicles. The purpose of this autonomous electric vehicle is to suppress human negligence when driving, besides being able to make it easier for the driver to travel without the need to drive it. Before all these things can operate properly, the autonomous electric vehicle needs a detection scheme that can distinguish objects, the segmentation method uses a thermal camera sensor based on Deep Learning that is trained with the required data set. This method uses a Fully Convolutional Network (FCN) with a Residual Network 34 (ResNet 34) architectural model with an image dimension of 640x512 pixels as its feature extraction. The advantage of ResNet 34 is that it is able to do quite a lot of dataset training even though the hardware used is not the most qualified. The design of this object detection system uses semantic segmentation, Neural Network, and Image Processing methods as input signals in the form of images, and a FLIR thermal camera which is like an eye for a vehicle which receives an input signal, the process from start to finish is processed using the Jetson AGX Xavier. The cability of semantic segmentation was tested offline with 40216 image frames, there are three categories, namely, good, sufficient, and not good. Which includes the good category as many as 23389 frames, 10706 frames is a sufficient category, and for not good category there are 6120 frames. The mean Intersection over Union (IoU) obtained at the time of this study was 78.56%.
KLASIFIKASI JENIS TAS PADA GAMBAR 360 DERAJAT (FISH EYE) DENGAN MENGGUNAKAN TENSORFLOW Pangemanan, Agnes Novi Anna; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.212

Abstract

Classification of objects is one of the studies that are currently being developed. The impact on the world of fashion is where women and men, teenagers to parents today cannot be separated from the bag as an addition to daily fashion. In the selection of bags, mistakes are often made, so as not to make a wrong choice, in this final project, I classify the types of bags, which is the method used to distinguish the characteristics of bags by type. The bag classification system based on type is a program that can identify a person's bag according to the type that has been trained and stored in the database of the program being run. Classification of bag types can be done in various ways, one of which is Deep Learning with the Convolutional Neural Network (CNN) method, CNN implementation using Tensorflow with the python programming language. This study was conducted using 5 classifications of bag type datasets totaling 6,720 images that have been trained with an image size of 180 x 180 using a 360o camera. It is hoped that this system is able to work well for classifying bag types in 360o (fish eye) image format. This study resulted in true detection rates of 55% and false detection of 45% where true detection is seen from the number of truths of accuracy in determining the output results, while false detection is the opposite of true detection from the number of 135 images that have been tested.
KLASIFIKASI WAJAH MANUSIA PADA GAMBAR 360 DERAJAT (FISH EYE) DENGAN MENGGUNAKAN TENSORFLOW Sutejo, Muhammad Fajar; Satyawan, Arief Suryadi; Siswanti, Sri Desy
Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (SENASTINDO) Vol. 4 (2022): Prosiding Seminar Nasional Sains Teknologi dan Inovasi Indonesia (Senastindo)
Publisher : Akademi Angkatan Udara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54706/senastindo.v4.2022.213

Abstract

Technology knows no boundaries, in fact it always shows new developments, one of which is classification in pictures. Human face classification is a method used to distinguish the characteristics of a person's facial pattern. The face classification system is an application that can find out a person's face according to the human face image that has been trained and stored in the machine's database. It is hoped that this application system can work well for classifying human faces in 360˚ image formats that have significant distortion Classification of human faces can be done in various ways, one of which is the Convolutional Neural Network (CNN) method using Tensorflow. This final project is carried out using 5 classifications of human face datasets totaling 6600 images that have been trained with an image size of 180 x 180 using a 360˚ camera and the Python programming language. The classification of human faces in 360˚ (fish eye) images was successfully carried out with a percentage of 65% true detection and 35% false detection from the total 135 images that have been tested. In further research, other deep learning methods can be used to obtain better classification accuracy
Deteksi Wajah Tersamar Menggunakan Metode VGGFace dan SVM Siswanti, Sri Desy; Puspita, Heni; Ubaya, Huda; Selly, Selly; Herdiana, Dina
REMIK: Riset dan E-Jurnal Manajemen Informatika Komputer Vol. 9 No. 2 (2025): Volume 9 Nomor 2 April 2025
Publisher : Politeknik Ganesha Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33395/remik.v9i2.14620

Abstract

Wajah adalah salah satu bagian dari manusia yang memiliki ciri-ciri berbeda. Teknologi pengenalan wajah merupakan suatu teknologi yang dapat mengidentifikasi atau memverifikasi seseorang dari sebuah gambar atau video. Teknologi pengenalan wajah bermanfaat untuk bidang keamanan, pengawasan, verifikasi identitas umum, sistem peradilan pidana, investigasi basis data gambar. Mungkin saja seorang DPO menggunakan penyamaran, baik secara sengaja maupun tidak sengaja, untuk menyembunyikan diri atau berpura-pura menjadi orang lain, misalnya menggunakan jenggot, kumis, dan gaya rambut yang diubah yang menyebabkan kebingungan dalam mengenali orang. Selain itu, aksesori penyamaran seperti wig, topi, syal, helm, kerudung, kacamata hitam, atau masker dapat membuat bagian wajah terlihat berbeda. Riasan tebal atau prosedur eksternal seperti operasi plastik juga dapat mengubah bentuk, tekstur, dan warna wajah, sehingga menyulitkan mengenali seseorang. Dalam makalah ini, mengusulkan sebuah algoritma pengenalan wajah tersamar,dimana algoritma ini mengubah arsitektur VGG pada tahap klasifikasi. Perubahan ini mencakup penambahan lapisan flatten yang disatukan dengan metode SVM. Tujuan dari modifikasi ini adalah untuk meningkatkan nilai akurasi dalam pengenalan wajah tersamar. Dalam penelitian ini memanfaatkan arsitektur VGG untuk ekstraksi fitur, SVM digunakan sebagai metode klasifikasi dalam pengenalan wajah. Sistem pengenalan wajah yang dikembangkan terdiri dari empat tahap utama: pengambilan data, pengolahan data, ekstraksi fitur, dan klasifikasi. Data wajah diambil secara langsung di depan kamera berupa wajah tanpa tersamar dan wajah tersamar dengan lima posisi wajah yaitu wajah menghadap ke kanan, ke kiri, ke depan,ke atas dan ke bawah. Sistem ini diimplementasikan menggunakan library Keras, Sklearn, dan Numpy untuk mengolah data. Untuk meningkatkan nilai akurasi diperlukan pengaturan parameter dari klasifikasi SVM yaitu Cost (C) dan gamma (ℽ). Hasil dari pengujian menunjukkan bahwa metode yang diterapkan dalam sistem pengenalan wajah tersamar ini menghasilkan nilai akurasi yang lebih baik dibandingkan dengan penelitian yang lain, walaupun masih ada beberapa kekurangan dari metode yang diterapkan dalam penelitian ini