Humpbacked Behavior Detection for Children with Autism Using Accelerometer in Smartphone

• Main Authors: Wei-Wei Yin, 殷維偉
• Other Authors: Hung-Huan Liu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/56258813994757056683

碩士 === 中原大學 === 電子工程研究所 === 100 === To establish a good interpersonal relationship face to face seems to be natural and easy for people. But that was challenges for autism children; when they want to interact with others. They sometimes don’t know how to properly communicate and to share. That because the barrier on the mental and physical development to such an extent that they learn language, social communication ability in slow progress, and therefore cause autism children often occurs stereotypical behavior and isolate themselves. The research target in this case study was an autistic student at a special education school, who often unconsciously became hunchbacked during group activities or when not talking to people. Currently, this student is already exhibiting slight scoliosis. We designed a system for hunchback detection and uploaded it to a smart phone. Combined with an assistive vest for operation, the system was installed on a smart phone, which was then turned on and placed in a pocket of the assistive vest. When the student wore the assistive vest, the system could determine whether the student’s back was hunched. Whenever the system assesses that hunchback behavior has occurred, the system will make an assessment and an alarm is emitted until the student leaves the hunchback state. The system also stores relevant data to establish goals for setting improvement plans or initiatives. Through field observations and fine tuning the system in response to feedback, we not only improved the hunchback behavior of the student, even achieving self-training effects, but also reduced the constant reminders and education workload of special education teachers.