Yew, Su Chiang
Date of Issue2015
School of Electrical and Electronic Engineering
Centre for Integrated Circuits and Systems
A highly integrated sports watch with an inbuilt accelerometer is capable of communicate with the computer through a CC1111 USB RF access point. The accelerometer is for measuring the accelerations point along the X, Y and Z axis. The data that collected from the accelerometer is from the orientation of the wearer's arm. Base on software, a hand gesture is implemented to control the mouse pointer using the wearer's arm acceleration data of the hand movement and the inferred hand position to perform on different gestures. This project design is to create a gesture based on window operation system that allowed the user to perform corresponding tasks with using the mouse pointer and gesture recognition in real time situation. To achieve mouse pointer control, continuous monitoring the orientation of the device which faces in three-dimensional systems is necessary. The recognition of the designed gesture is dependent on the collected acceleration data of the three axes of the accelerometer from the device. The collected acceleration data within a data buffer must not exceed the threshold value that is set by the user, and this data will save into a single file to recognize as a gesture. The next gesture classification, such as learning and recognition is to use multiple layer perceptron (MLP) to learn the pattern or action independently and automatically. When the user-defined a direction of the wearer's arm, after completed the learning and recognition process, it must be able to execute the recognizing gesture with the existing data. Lastly, the selected software in this design must be capable of maintaining the executed gesture action profiles by the user, without interference or stop the wireless connection while interacting with the computer.
DRNTU::Engineering::Computer science and engineering::Software::Software engineering
Final Year Project (FYP)
Nanyang Technological University