Analyzing Video Sequences of Multiple Humans door Jun Ohya, Akira Utsumi & Junji Yamato