Deteksi Dan Prediksi Trajektori Objek Bergerak Dengan Omni-Vision Menggunakan Pso-Nn Dan Interpolasi Polynomial

Setyawan, Novendra; Mardiyah, Nur Alif; Hidayat, Khusnul

doi:10.24269/mtkind.v13i1.1691

Cited by 4 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Sistem kontrol dapat bekerja secara otomatis dalam banyak cara [6]. Pada sistem prediksi dengan interpolasi polynomial akurasi prediksi didapatkan 20 % yang membutuhkan peningkatan sistem metode interpolasi agar akurasi meningkat [7]. Metodenya adalah dengan mengubah citra dari RGB ke skala abu-abu [8].…”

Section: Pendahuluanunclassified

Sistem Deteksi dan Pelacakan Bola dengan Metode Hough circle Transform Menggunakan Kamera Omnidirectional pada Robot Sepak Bola Beroda

Putra

Puriyanto

2022

Buletin Ilmiah Sarjana Teknik Elektro

View full text Add to dashboard Cite

This research develops a wheeled soccer robot or KRSBI-B in detecting and tracking the ball using an omnidirectional camera. The disadvantage of the previous robot is that the robot is less accurate in detecting the ball because the robot's visibility to the ball and the intensity of light on the field are different. With the Hough circle transform method and color filtering, this study conducted several tests such as detecting the ball if it was covered by an opposing robot, then getting the coordinates of the robot's distance from the ball, and testing different light intensities in the room. The results achieved are that the system can detect and track the ball with the results of object detection between the robot and the ball from the front, the result is the difference in coordinates (x, y) on pixels and coordinates (x, y) references with an overall average error value of 0, 14%. Detection of objects between the robot and the ball from all directions with 24 times of testing the results obtained accuracy and precision of 100% and a miss rate of 0%, then on testing changes in lighting intensity with 4 different lighting tests and different distances, accuracy and precision results are obtained. by 100% with a miss rate of 0%. Detection of the ball if it is covered by the robot, the system is able to detect the ball up to 75% of the range of objects on the side of the ball and shows an average Peak Signal to Noise Ratio (PSNR) image quality of 39.6% with good image quality results indicated by a high PSNR value tall. Penelitian ini mengembangkan robot sepak bola beroda atau KRSBI-B dalam mendeteksi dan melacak bola menggunakan kamera omnidirectional. Kekurangan pada robot sebelumnya yaitu kurang akuratnya robot dalam mendeteksi bola karena jarak pandang robot terhadap bola dan intensitas cahaya di lapangan yang berbeda-beda. Dengan metode Hough circle transform dan colour filtering penelitian ini melakukan beberapa pengujian seperti mendeteksi bola jika tertutup oleh robot lawan, kemudian mendapatkan koordinat jarak robot terhadap bola, dan pengujian intensitas cahaya yang berbeda-beda di dalam ruangan. Hasil yang dicapai yaitu sistem dapat mendeteksi dan melacak bola dengan hasil dari deteksi objek antara robot dan bola dari arah depan didapatkan hasil selisih koordinat (x, y) pada pixel dan koordinat (x, y) referensi dengan keseluruhan nilai error rata-rata 0,14%. Deteksi objek antara robot dan bola dari segala arah dengan 24 kali pengujian didapatkan hasil akurasi dan presisi sebesar 100% dan miss rate 0%, kemudian pada pengujian perubahan intensitas pencahayaan dengan 4 kali pengujian pencahayaan yang berbeda dan jarak yang berbeda di dapatkan hasil akurasi dan presisi sebesar 100% dengan miss rate 0%. Pendeteksian bola jika tertutup oleh robot, sistem mampu mendeteksi bola hingga rentang objek sisi bola terlihat 75% dan menunjukkan hasil kualitas citra Peak Signal to Noise Ratio (PSNR) rata-rata 39,6% dengan hasil kualitas citra yang baik ditunjukkan dengan nilai PSNR yang tinggi.

show abstract

Section: Pendahuluanunclassified

Sistem Deteksi dan Pelacakan Bola dengan Metode Hough circle Transform Menggunakan Kamera Omnidirectional pada Robot Sepak Bola Beroda

Putra

Puriyanto

2022

Buletin Ilmiah Sarjana Teknik Elektro

View full text Add to dashboard Cite

show abstract

“…Permanent magnet DC motors (PMDCM) has been used in many application such as automotive, computer zapplication, robotic [1][2] [3][ [4], and industrial application [5][6] [7] [8][6] [9]. Its advantages over other conventional motors are better speed and torque characteristics, better dynamic response, high efficiency, no need for excitation current, no noise operation, high weight to torque ratio, and relatively low cost [10].…”

Section: Introductionmentioning

confidence: 99%

Hybrid Frequency and Period Based for Angular Speed Measurement of DC Motor Using Kalman Filter

Setyawan¹,

Cahyadi

Hakim³

et al. 2022

KINETIK

Self Cite

View full text Add to dashboard Cite

The Incremental Rotary Encoder have been widely used to measure the angular speed of electrical drive such as Permanent Magnet Direct Current Motor (PMDCM). Nevertheless, speed measurement of PMDCM from the encoder signals can be subject to errors in some special condition such as in low resolution encoder. There are two main methods to measure the angular speed of PMDCM through encoder signal such as frequency-based and period-based wich has its own properties. Hence in this reseach aimed to improve the angular speed measurement with hybridization of frequency and period-based measurement. The Hybrid method is defined as paralleling the period and frequency then estimated the angular speed using sensor fusion with Kalman Filter. The experiment is doing by comparing of all method to get the best way in measuring. From the experimental showed that the Kalman filter parameter was fine tuned that resulting the sensor fusion or the mixed measurement between the frequency-based and the period based measure the angular speed accurately.

show abstract

“…Recently, we can find so many works regarding time series analysis, using different kinds of approaches. Some were using the fuzzy time series approach [3][4][5][6][7], while some others were using hybrid approach [8][9][10][11][12], combining two or more methods to forecast the time series data. In this paper, we usea different approach, using backpropagation neural networks to forecast Jakarta Stock Exchange (JKSE) data.…”

Section: Introductionmentioning

confidence: 99%

Backpropagation Neural Networks Implementation for JKSE Forecasting

Hansun*

2019

IJRTE

View full text Add to dashboard Cite

Neural networks is a type of soft computing methods that widely has been used and implemented in many fields, including time series analysis. One of the goals of time series analysis is to predict future data value.In this study, we try to implement another approach using the backpropagation neural networks method to forecast the Jakarta Stock Exchange (JKSE) composite index data, which is one of the stock market change indicators in Indonesia.The study then is continued by calculating the accuracy and robustness levels of Backpropagation NN in forecasting JKSE data. The experimental result on the case taken shows an encouraging and promising result.

show abstract

Deteksi Dan Prediksi Trajektori Objek Bergerak Dengan Omni-Vision Menggunakan Pso-Nn Dan Interpolasi Polynomial

Cited by 4 publications

References 6 publications

Sistem Deteksi dan Pelacakan Bola dengan Metode Hough circle Transform Menggunakan Kamera Omnidirectional pada Robot Sepak Bola Beroda

Sistem Deteksi dan Pelacakan Bola dengan Metode Hough circle Transform Menggunakan Kamera Omnidirectional pada Robot Sepak Bola Beroda

Hybrid Frequency and Period Based for Angular Speed Measurement of DC Motor Using Kalman Filter

Backpropagation Neural Networks Implementation for JKSE Forecasting

Contact Info

Product

Resources

About