Skripsi

PERANCANGAN SISTEM PENGENALAN OBJEK PADA HUMANOID ROBOT SECARA REAL TIME BERBASIS DEEP LEARNING.

---

Penilaian

---

The development of today's robots is very fast, there are even robots that have shapes and abilities that are very similar to humans, such as interacting with humans, as well as the ability to recognize objects around the robot. Recognition of objects carried out by robots usually uses the help of a camera. However, at this time the object recognition has not been realized in real time on humanoid robots. So, in this study, an object recognition system will be developed around the robot, especially in the control system and robotics laboratory at the Unsri Indralaya campus. The algorithm used in this research is using the Convolutional Neural Network (CNN) algorithm. In this study, the objects detected were divided into 12 classes, namely bags, books, bottles, chairs, doors, eggs, laptops, glasses, people, tables, blackboards, and windows. The data used for object recognition is primary data taken at the Control and Robotics Engineering laboratory, which in this experiment uses 15,000 training data and 4,000 as test data. There were two experiments in this study, namely object recognition at a close distance of one meter and object recognition at a distance of two meters. The results showed that the accuracy in the simulation experiment was 60% and 100% at epoch 200 and 500. Then the accuracy in real time conditions was 60% at close range and 90% at long distance. This shows that the Convolutional Neural Network (CNN) algorithm using the ssd_mobilenet_v1_pets model has good performance in object recognition systems around humanoid robots

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2407001566	T140731	T1407312023	Central Library (Referens)	Available

Detail Information

Series Title

-

Call Number

T1407312023

Publisher

Indralaya : Prodi Teknik Elektro, Fakultas Teknik Universitas SriwijayaC., 2023

Collation

xii, 53 hlm.; ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

629.807

Content Type

Text

Media Type

unmediated

Carrier Type

-

Edition

-

Subject(s)

Robot
Prodi Teknik Elektro

Specific Detail Info

-

Statement of Responsibility

MURZ

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment