Haiqal Ramanizar Al Fajri, . (2022) SISTEM PENGENALAN GERAK BAHASA ISYARAT DENGAN COLORED MOTION HISTORY IMAGE DAN CONVOLUTIONAL NEURAL NETWORK. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (15kB) |
|
Text
AWAL.pdf Download (530kB) |
|
Text
BAB 1.pdf Download (85kB) |
|
Text
BAB 2.pdf Restricted to Repository UPNVJ Only Download (425kB) |
|
Text
BAB 3.pdf Restricted to Repository UPNVJ Only Download (142kB) |
|
Text
BAB 4.pdf Restricted to Repository UPNVJ Only Download (1MB) |
|
Text
BAB 5.pdf Download (71kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (92kB) |
|
Text
RIWAYAT HIDUP.pdf Restricted to Repository UPNVJ Only Download (79kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (7MB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (27MB) |
|
Text
ARTIKEL KI.pdf Restricted to Repository staff only Download (468kB) |
Abstract
This research was conducted to create a sign language recognition system that can be used to recognize sign language movements in the Indonesian Sign Language (BISINDO) system. Sign language is a method of communicating for deaf people to understand the meaning and information received and convey desires and emotions using the help of hands, gestures, lips, and facial expressions. In this research, the author uses the Convolutional Neural Network (CNN) method to perform the motion recognition process. In addition, the author uses the Colored Motion History Image (Colored MHI) method to represent motion from video into one image. The Colored MHI method performs color changes made by the Motion History Image (MHI), which generally uses a grayscale into RGB color format. The data was obtained through video shooting by the author on 15 subjects with 5 movement classes and resulted in a total of 450 data. The video data that has been obtained is cropped and then converted into a single image using the Colored MHI method. The results of making the CNN model with training data are tested with test data that has passed the Colored MHI stage and its performance will be seen through its accuracy and loss values. The results of this research indicate that the CNN and Colored MHI methods can recognize sign language gestures quite well. The accuracy and loss obtained are 0.8533 and 0.4741.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil: 1810511007] [Pembimbing 1: Jayanta] [Pembimbing 2: Bambang Tri Wahyono] [Penguji 1: Yuni Widiastiwi] [Penguji 2: Nurul Chamidah] |
| Uncontrolled Keywords: | Convolutional Neural Network, Motion History Image, sign language, BISINDO. |
| Subjects: | Q Science > Q Science (General) Q Science > QA Mathematics Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > T Technology (General) |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | Haiqal Ramanizar Al Fajri |
| Date Deposited: | 19 Aug 2022 06:49 |
| Last Modified: | 19 Aug 2022 06:49 |
| URI: | http://repository.upnvj.ac.id/id/eprint/19673 |
Actions (login required)
 |
View Item |