Developing a Robotic Handwriting Recognition System Using Neural Network
By
Du Hui | Student |
Chang Wei Hong | Student |
Zhang Qian | Student |
Bryan Chen | Student |
Foo Tuan Hoe Edwin | Lecturer |
|
|
Nanyang Polytechnic | Singapore |
The Challenge
To develop a system to help the people with writing difficulty to write properly by interpreting the numbers they have written. This system can also be used to help train kids to write properly and arouse their interest in learning to write properly.
It can also serve as helping tools for those with writing difficulties in writing. Thus, the challenge is to provide a drawing box in LabView for the user to draw free-hand and for the system to be able to recognize the numbers using of neural network and interpret the number correctly. The number, will be written by controlling the robotic arm in 7-segment format.
The Solution:
Lab View provides a graph box for user to draw. Collecting the coordinates of the point of the drawing, we are able to match them with pre-defined coordinates and identify the digit drawn. We will use the PC to control the robotic arm to draw on a sheet of paper. The instruction packets of the command will be sent through the RS232 cable and translated to RS485 by the CM-2 Dynamixel controller board that controls the robotic arm.
ATTACHED: ORIGINAL REPORT
