13 research outputs found
Pendeteksian Objek Dalam Jaringan Sensor Nirkabel Multimedia
No full textObject detection applications can be used to perform surveillance functions on car
traffic. A car can be identified by its type and license number (plate). Meanwhile,
the reliability of multimedia sensor networks for car traffic monitoring can be
considered based on the number of multimedia sensors used and the positioning of
the sensors. This study uses a camera sensor with Phase Detection Auto Focus
(PDAF) technology, similar to a pair of right and left eyes when looking at an
object. This research discusses car detection by multimedia sensors in a wireless
multimedia sensor network used for car traffic monitoring. The study used 2 car
objects to detect the type and license plate. However, the research focuses on
detection by proposing a car and license plate detection model called Faster RCNN
IFY, where its detection performance is compared to the YOLOv5 model and
Commercial ALPR applications. Based on the detection, the disparity value
obtained at a maximum distance of 50 m is 6.20 x 106 pix/mm, while the depth
image value at the maximum disparity value is 16.88 x 109 mm3/pix. Then, in the
proposed model, with a testing dataset of 40 images using the Faster R-CNN IFY
model, the average value of car plate detection accuracy is quite good, namely
80.67%. Although these results are not as accurate as when the same testing dataset
is applied to the Commercial ALPR application, which is 87.9%. The more training
datasets, the better the accuracy of the detection results. Therefore, in the Faster
R-CNN IFY model with a training dataset of 240 images, the average accuracy of
the car detection results is only 52.6%. Whereas in the YOLOv5 model, with a total
training dataset of 1500 images, an increase in the average value of the accuracy
of the detection results was obtained to be 87.1%. In the end, in this study the Faster
R-CNN IFY model was superior in terms of speed compared to the Commercial
ALPR and YOLOv5 models. The average time needed to test 40 car images using
the YOLOv5 model is 111.43 msec. Meanwhile, applying the Faster R-CNN IFY
model only takes 11.86 msec. Likewise, when detecting 40 car plate images using
Commercial ALPR, it takes an average time of 288.76 msec, while using the Faster
R-CNN IFY model it only takes 11.86 msec. In addition, this model is also able to
simultaneously detect 2 object classes in a car, namely the car class and license
plate class. Meanwhile, Commercial ALPR and YOLOv5 can only detect 1 class of
objects in a car.77 HalamanDisertasi Dokto
Vehicle detection system based on shape, color, and time-motion
Get PDFVehicle detection application can assist in-vehicle surveillance functions and have implications for various fields. A vehicle can be identified through the license number attached to its license plate, the color and its shape. Vehicle detection can make use of multimedia sensors so that the design and detection performances can be optimal. Sensor performances are influenced by factors such as the number of multimedia sensors, sensor placement, sensor positioning, and schemes in case of system failure. This study makes use of multimedia sensors with cameras equipped by a phase detection auto focus (PDAF) technology which is like a pair of eyes to see an object. This study analyses 134 vehicles with number detection and various colors to see the effect on the detection and recognition processes. The cars were passed through the camera 10 times at a speed of 10-15 km/hour with various camera distances and positions. Various values and depths of the images were generated. The farther the distance the higher the disparity values. For maximum distance of 50 m, disparity is 6.20×106 and image depth is 16.88×109. Vehicle color influences detection with orange has the best accuracy, but the gray has the largest path error value
PPTTG Bandrek Jahe Merah Adelia Di Desa Lau Bekeri Kecamatan Kutalimbaru Kabupaten Deli Serdang Provinsi Sumatera Utara
Full text linkPengabdian Penerapan Teknologi Tepat Guna Kepada Masyarakat (PPTTG) di UMKM Bandrek Jahe Merah Adelia Di Desa Lau Bekeri Kecamatan Kutalimbaru Kabupaten Deli Serdang Provinsi Sumatera Utara dilakukan dengan tujuan untuk mengatasi permasalahan yang dihadapi mitra dalam melakukan pengeringan bubuk jahe merah, kunyit dan temulawak yang mengkristal akibat kegagalan produksi. Selain itu proses produksi yang masih dilakukan secara manual hanya menghasilkan kapasitas produksi yang terbatas. Sehingga kadang tidak dapat memenuhi permintaan/kebutuhan konsumen dalam porsi besar. Solusi dari permasalahan tersebut yaitu memberikan peralatan berupa smart oven pengatur suhu otomatis untuk pengering serbuk bandrek yang merupakan modifikasi dari Hasil Penelitian Produk Terapan tim PPTTG pada tahun 2019 dan mesin pengaduk air pati jahe merah/kunyit/temulawak. Sehubungan dengan masa pandemi korona yang mengharuskan semua pihak untuk Work From Home (WFH), maka metode yang paling bijaksana digunakan dalam PPTTG ini, adalah dengan mengumpulkan data/informasi secara daring. Selanjutnya dianalisis secara deskriptif, diimplementasikan dan dilakukan evaluasi terhadap program secara keseluruhan. Kegiatan PPTTG ini akan dilakukan dalam 4 tahapan, dimulai dari tahap persiapan dan pengumpulan data, tahap kedua melakukan desain/merancang dan memodifikasi smart oven sesuai dengan kebutuhan mitra. Tahap ketiga melakukan desain/merancang mesin pengaduk air pati jahe merah/kunyit/temulawak untuk memudahkan proses pembuatan bandrek jahe merah dan bandrek jahe merah plus kunyit temulawak. Tahap keempat review tim pemantau PPTTG dan evaluasi kembali oleh Tim PPTTG untuk melihat kesesuaian outcomes. Luaran yang akan dihasilkan dari kegiatan PPTTG ini, adalah berupa produk/barang, publikasi pada media massa cetak/daring, publikasi berupa video yang bisa diakses secara daring, publikasi pada jurnal nasional terindeks SINTA dan pencatatan hak cipta (HKI)
System of smart detection and control to electrical energy for saving of electrical energy consumption
Full text linkPublic campus has a mandate to saving of electrical energy. Electrical energy consumption is often wasteful in building. There is tendency wasteful by user. Electronic equipment is often still turn on at idle time. Only a few students want to turn off the equipment and shut down the computer. Saving of electrical energy is not only at idle time but it can be improved into operational hour. It is not depending on idle time or operational hours, but depends on human presence. Implementation of electrical energy saving has to be supported by frugal behavior and equipment technology. In this study, we name system of smart detection and control to electrical energy (Sisdece). This system is consist of hardware and software. Hardware applies passive infrared sensor (PIR) sensor, wireless sensor network (WSN), microcontroller ESP32, access point, relay. Software use C++, hypertext preprocessor (PHP), hypertext markup language (HTML) and android studio. Result of measurement has been done in a month during November 2020. Average of energy saved is 12.51 kWh and total of electrical energy is 105.86 kWh. Comparison of energy saved to electrical energy is 11.81%. This is a significant reduction to electrical bill. The result is expected as benchmark of electrical energy management in Politeknik Negeri Medan (POLMED)
PKM Usaha Sablon Nyablon.061 di Tanah Enam Ratus Marelan Medan
Full text linkProgram Kemitraan Masyarakat (PKM) ini bertujuan untuk mengatasi permasalahan mitra yaitu Usaha Sablon Nyablon.061 dalam meningkatkan produksi dan pemasaran hasil usahanya. Sedangkan target khusus kegiatan PKM ini adalah memberikan alat sablon 5 in 1 untuk meningkatkan produksi melalui diversifikasi produk yang disertai dengan pelatihan penggunaan alat tersebut, peningkatan kemampuan mendesain logo dan merk melalui pelatihan desain grafis, dan pembuatan album foto produk guna menguatkan metode pemasaran hasil produksinya. Metode yang digunakan dalam PKM ini adalah pengumpulan data/informasi melalui survey dan wawancara, kemudian data dianalisis secara deskriptif, diimplementasikan, dan dievaluasi secara keseluruhan. Kegiatan PKM ini dilakukan dalam 3 tahapan, yaitu tahap persiapan melalukan survey dan wawancara untuk mengumpulkan data dan informasi tentang mitra, tahap pelaksanaan melakukan pemberian alat produksi dan pelatihan, tahap evaluasi melakukan review dan evaluasi oleh tim PKM dan pemantau PKM untuk melihat kesesuaian outcomes
PKM INSTALASI PANEL SURYA SEBAGAI SUMBER LISTRIK DI TAMBAK RAKYAT PALUH MERBAU DESA TANJUNG REJO KECAMATAN PERCUT SEI TUAN KABUPATEN DELI SERDANG SUMATERA UTARA
Full text linkIndonesia merupakan negara pengekspor ikan dunia. Pemerintah melalui Kementerian Perikanan dan Kelautan mendukung pengembangan budidaya ikan yang mandiri, berdaya saing dan berkelanjutan. Tambak rakyat paluh merbau merupakan salah satu usaha tambak milik masyarakat paluh merbau di Desa Tanjung Rejo. Tambak ini memiliki memiliki kincir sebagai pensuplai oksigen kedalam kolam. Fungsi kincir juga sebagai pembersih area permukaan air dan dasar air sehingga menciptakan arus air yang stabil. Untuk menggerakkan kincir digunakan listrik dari PLN. Kendala yang dihadapi oleh tambak ini adalah biaya pakan yang mahal dan biaya listrik yang juga mahal. Biaya pakan tidak bisa dikurangi karena terkait dengan produksi ikan. Namun biaya penggunaan token listrik dapat dihemat dengan penggunaan sumber listrik tenaga surya. Melalui kegiatan PKM ini akan dibangun instalasi panel surya dengan sistem hybrid sebagai sumber listrik untuk menggerakkan kincir di tambak. Diharapkan dengan kegiatan ini dapat menghemat penggunaan token listrik di tambak rakyat paluh merbau sehingga bisa mengurangi biaya produksi
Implementation of the RS232 communication trainer using computers and the ATMEGA microcontroller for interface engineering Courses
No full textDesign of environmental detector system application aims to promote awareness of pollution on campus
Full text linkPoliteknik Negeri Medan (POLMED) was involved in the UI GreenMetric world rankings. The UI GreenMetric committee assessed green campus activities and environmental sustainability. The UI GreenMetric aims to raise awareness about sustainable campus greening, and social impacts of these endeavors. Based on the concept, an environmental detection system (EDS) was developed using internet of things (IoT) technology. The EDS can detect and monitor environmental parameters remotely such as carbon dioxide (CO2), noise levels, light intensity, air temperature, relative humidity, and dust particle density in real-time via the internet. Measurements of environmental parameters were conducted at one location in POLMED. The average CO2 level was 485 ppm. The average noise level was 53.40 dB. The average light intensity was 129 lux. The average air temperature was 26.60 °C. The average of relative humidity was 63.8% RH. The average of PM2.5 dust particle densities was 23 µg/m3. The average of PM10 dust particle densities was 29 µg/m3. Based on these results, the air quality has begun to be polluted because this value is already above the threshold clean quality air set by the Government of the Republic of Indonesia (310–330 ppm)
Urban Planning with the Combined Method of Perception-Driven Joint Learning Approach (PeDJoLA) and Geographic Information Systems (GIS) Model for Disaster Mitigation
No full text
