The investigators plan to conduct a first-stage experiment by recruiting ten subjects, including five PD patients and five non-PD patients. In the beginning, the information collected by the portable motion detector is used to compare the difference in the activities performed by PD and non-PD patients. Besides, the algorithm will be developed to identify the gait patterns of PD patients. Once the system detects the abnormal gaits, such as shuffling and festinating steps or freezing of gait, auditory cues will be given to the PD patients and caregivers by their sides. It can help the patient to maintain normal gait and improve the effectiveness of rehabilitation, as well as preventing falls in daily activities. Moreover, in case of accident falls, the real-time fall detection mechanism alerts the nearest caregiver for instant support and delivers this information to the remote family members and medical personnel.

After that, a second-stage experiment will be carried out by recruiting another five PD patients. By comparing the gait patterns identified by the system against those identified manually by the staff, the performance of the proposed system on fall prevention will be examined. For the patients and their family members, the developed system enhances the patients' safety in their daily activities. As for the medical personnel, it serves as an affordable tool that benefits the rehabilitation of PD patients in an easy manner.

Inclusion Criteria:

• Clinical diagnosis of Parkinson's Disease

Exclusion Criteria:

• Cardiac disease
• Mobility disability