Enhancing Visual Recognition for Door Status Identification in AAL Robots via Machine Learning

In recent years, there has been a growing trend to increase the inclusion of robotic platforms in Ambient Assisted Living (AAL) environments as assistive robots, the main goal of which is to extend independent living and improve quality of life of the elderly. The use of robotic platforms plays an increasingly critical role in the daily life, especially when incorporated into Ambient Assisted Living environments, as they can help transfer objects, monitor people's activities, identify emergencies etc. by exploiting robotic capabilities such as autonomous navigation, manipulation or perception. With the role of the robotic platform being to assist the tenant in his daily activities, the robot must be able to visualize and interact with objects through machine vision and machine learning, based on their morphology; at the same time it must be able to interact and recognize the home environment. In this paper we propose an efficient method for visual recognition of door status from algorithms running on assistive robots.