Contextual evaluation of the impact of dynamic urban window view content on view satisfaction

Abstract

Movement is a preferred attribute in window views, but identifying the ideal amount of movement for different space types remains challenging. This study examined the impact of movement in urban window views on view satisfaction in four spaces: office, living room, restaurant and classroom. Fifty dynamic urban window views were recorded at 30 frames per second, and movement was quantified by examining the percentage of pixels that exhibited intensity changes between successive video frames. Window view compositional ratios were determined through semantic segmentation using a pretrained DeepLabv3+ model. Fifty participants evaluated the movement and view satisfaction in a virtual reality survey. Movement was sufficient when average change in the view after every 1/30 s ranged between 2 % to 8 % during the viewing period. Correlation analysis revealed that the strongest indicators of excessive, insufficient and sufficient movement are Ratio of Human-associated Dynamic Objects (RHDO), Building Ratio (BR), and Greenery Ratio (GR), respectively. It was established that occupants can tolerate higher movement and RHDO in restaurants and offices than in living rooms and classrooms. These insights can inform the allocation of space in mixed-use buildings. The movement quantification method can also be incorporated into other dynamic window view evaluation frameworks.