Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Nov 6, 2012
I constructed a novel system that uses vertical Kinect sensor for people counting, where the depth information was used to remove the effect of the appearance variation. Since the head is always closer to the Kinect sensor than other parts of the body, people counting task equals to find the suitable local minimum regions. According to the particularity of the depth map, I proposed a novel unsupervised water filling method that could find these regions with the property of robustness, locality and scale-invariance. The water filling algorithm just simulated the process of the raindrops drop down from the sky to the land and move to the nearest hollow region. I took the depth map as a land and dropped the rain to it, the region gathered with drops will be classified as a hollow, which means the head of people.(AVSS 2012 paper, accepted as oral), you can download the the ppt: https://drive.google.com/file/d/0B6Si... paper: https://drive.google.com/file/d/0B6Si...