Learned Hand Gesture Classification through Synthetically Generated Training Samples

Abstract:

We describe the design and working of a robotic device that can be used for the rescue of children trapped in borewells. The design is inspired by taking into account the reasons for failure of previously existing methods. A key feature of the device is its ability to lift a child with the help of a safety cage, instead of a grasping mechanism. The device has the capacity to shrink the size to lower into the borewell, thereby not affecting the wall. The joint angles have a facility of rotating 360 degrees, enabling the device to find a gap in any direction, without hurting the child. The radial movement of the device also enables it to function for a borewell ranging from 25.4 centimeters to 45.72 centimeters in diameter. The suitability of the proposed system is demonstrated by several experiments to test reliability of the developed prototype.

Download paper here

Recommended citation: Pattery, John Jose, Niveditha Kalavakonda, Jitu Verghese Kurian, Lekha Mohan, Jippu Jacob. “Design and Development of Robot for Borewell Rescue”, Proceedings of the 14th IFToMM World Congress ; 2015 Oct 25-30; Taipei, Taiwan. doi:10.6567/IFToMM.14TH.WC.OS13.103