0610-nsf-eldertechnology-grant
Back projection of two-dimensional image plane silhouettes in three-dimensional space for voxel person construction.
Photo:
Download larger image
Activity sequence, same as figure 4, which consists of 3650 frames (approximately 12 minutes using a capture rate of 5 fps). The output of reasoning about the state of voxel person (level one of fuzzy inference) is plotted over the time/frame domain. Labels indicate where the second level of fuzzy inference classified a fall. This sequence contained three falls, all of which were correctly recognized.
Photo:
Download larger image
Examples of spine tracking interface layout that is shown to research participants.
Photo:
Download larger image