Department of Computer Engineering
MS THESIS PRESENTATION
Recognition of Occupational Therapy Exercises for Cerebral Palsy
Mehmet Faruk Ongun
(Supervisor: Prof. Dr. Uğur Güdükbay)
Computer Engineering Department
Virtual rehabilitation methods based on depth cameras for occupational therapy is one of the popular methods for approaching patients with cerebral palsy. There are some obstacles to overcome when designing an exercise recognition system for this specific purpose. This thesis proposes a combination of solutions targeting these obstacles. The proposed solution works for upper extremity functional exercises. We first aim to eliminate the gestures that are not defined in our exercise set and manage it using a method called universal negative model. Then, we determine the compensation mistakes, that are done frequently by the patients during the exercise sessions, and restrictions for classifying these mistakes. We use specific negative models based on feature thresholding and train them specifically for each exercise to distinguish the determined compensation mistakes. We conducted various tests using our method in a laboratory environment under the supervision of an occupational therapist and presented the successful results.
DATE: 14 September 2018, Friday @ 10:00