Automated estimation of elder activity levels from anonymized video data

Significant declines in quality of life for elders in assisted living communities are typically triggered by health events. Given the necessary information, such events can often be predicted, and thus, be avoided or reduced in severity. Statistics on activities of daily living and activity level over an extended period of time provide important data for functional assessment and health prediction. However, persistent activity monitoring and continuous collection of this type of data is extremely labor-intensive, time-consuming, and costly. In this work, we propose a method for automated estimation of activity levels based on silhouettes segmented from video data, and subsequent extraction of higher order information from the silhouettes. By building a regression model from this higher order information, our system can automatically estimate elder activity levels.