Assisted child-minding based on visual activity monitoring in a home camera surveillance system

• Main Author: Chng, Carolyn Jun Er
• Other Authors: Cham Tat Jen
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/65594
• Institution: Nanyang Technological University

This report involves discovering ways to prevent accidents from happening at home for young children. Although the childminders claimed that the child are within their reach when they are at home, there is no standard measurement to define ‘within their reach’. As the injury rates and pattern fluctuate from one country to another, regardless of regions, developing or non-developing, home accidents still happen. Hence, the major themes derived from literature reviews are home environment, lack of parental supervision, injuries happening in the inequality of gender as well as home safety. Even though the rate of growth in children is something not within human control, the facts that infants facing home accidents are worth the effort to make the home environment a safer place for their children through the mentioned precautions. Moreover, this project aims to track the position of the child by developing an interactive and semi-automated system. The GUI is a simplicity based design that almost fulfilled the Shneiderman’s ‘Eight Golden Rules of Interface Design’ such that non-tech savvy childminders can used it. Additionally, the application allows childminder to track and monitor the child location along with the identifying of home’s restricted zone. A triggered alarm from the application is implemented to notify the childminders when the child enters the restricted zone. With this application, people will be able to focus on their tasks at hand and also assist in taking care of the child. The main objective of the project are fulfilled though the accuracy of application are dependable on the lighting condition of images as the result varies with the change of lighting condition. Recommendation such as improving the algorithm and implementing SMS function in the application are suggested. This concludes the functions that the application can achieve.