Michael Hild

Learn More
This paper describes a view-based localization method using support vector machines in outdoor environments. We have been developing a two-phase vision-based navigation method. In the training phase, the robot acquires image sequences along the desired route and automatically learns the route visually. In the subsequent autonomous navigation phase, the(More)
This paper describes a view-based outdoor navigation method. In the method, a user first guides a robot along a route. During this guided movement, the robot learns a sequence of images and a rough geometry of the route. The robot then moves autonomously along the route with localizing itself based on the comparison between the learned images and input(More)
This paper describes a panoramic view-based navigation in outdoor environments. We have been developing a two-phase navigation method. In the training phase, the robot acquires image sequences along the desired route and automatically learns the route visually. In the subsequent autonomous navigation phase, the robot moves by localizing itself by comparing(More)
  • Michael Hild
  • Proceedings of the 17th International Conference…
  • 2004
We propose a simple method for human gait recognition from monocular color image sequences which recovers the approximate 3D motion trajectory. We provide concrete evidence that object-to-image height ratios can be used for estimating 3D motion trajectories and motion velocities with sufficient accuracy for making judgments about the states of motion. We(More)
We propose an object recognition method in which the identity of objects are determined by observing the finger pointing action of persons. The system determines the head and hands regions, tracks them in real time, and verifies them through an analysis of 3-D data points. Then it determines the pointing direction and intersects it with an environment(More)
This paper describes an imaging system for the acquisition of physically disparate images, such as visible and infrared wavelength range images, together with an image registration algorithm. The algorithm makes use of 3D scene data obtained from stereo and integrates local and global registration methods in order to obtain registrations that have good(More)