Student Projects
Please complete the following information
Please complete the following information (영문으로 작성바랍니다.)
Contact Information
University and Department: Mechatronics engineering dept of Korea Polytechnic University
Team Members: Byung-Wook Choi, Myung-Jin Back, Soon-Shin Kang, Kyu-Dae Shim, Jung-Min Um
Faculty Advisors: Dr. Dae-Jin Kang
Primary Email Address:djkang@kpu.ac.kr
Primary Telephone Number (include area and country code):
82-10-2220-5698, 82-31-8041-0453
Project Information
Project Title: Sign Language Translator
List all parts (hardware, software, etc.) you used to design and complete your project:
- Personal computer
- PCI-6221
- Microphone & Amplifier
- Voice recognition module
- Webcam
- Speaker
- LabVIEW
- Vision module
Describe the challenge your project is trying to solve.
We tried to solve the understanding inconvenience of the sign language between handicapped person like dumbness and us.
The sign language is a translation tool of persons who can not speak.
But there’s translation problem between person who can’t use the sign language and handicapped person at public place, for example, the bank, bus terminal, municipal office etc.
We can solve this problem using Vision as followings.
- Some handicapped person like dumbness visits at bank
- Banker asks something for example ID card
- Voice recognition module send a matched signal which is prepromised to DAQ board
- Personal computer displays sign language as matched signal to handicapped person
- Handicapped person says something using sign language
- Taking a picture of sign language of handicapped person using webcam continuously
- Compare the picture of database which already be made
- Translate the sign language motion to normal language
- Banker understands sign language using speaker
- Repeat upper procedure
Describe how you addressed the challenge through your project.
We’re faced with several difficulties at this project. Firstly, there’s a performance limitation on voice recognition module. We assumed a situation as bank affairs. Therefore, we designed translation program within 8 words. Secondly, the recognizing rate of vision was low at first time. We can solve this problem to control the light intensity, back ground & parameter tuning.
If we will use more high performance voice recognition module, we will be able to succeed at much more conditions.
We made it using Korean sign language now, but it could be changed easily to another country sign language.
We believe this results will give a chance to remove the barrier between handicapped person like dumbness and us.
Related photo (이메일에 파일로 별도 첨부 바랍니다.)
Related video (이메일에 파일로 별도 첨부 바랍니다.)
http://www.youtube.com/watch_private?v=RPu58tKfWv8&sharing_token=vAoEwKEQXpOVU_oNxHUlnw
We have small problem to generally open this video at Youtube because patent except NI
If you need any permit to open this video, please contact to us(djkang@kpu.ac.kr)
I will send all material including video to Mr. Seil Park of NI Korea.
