A Human-Hand Detection System, Apparatus and a Method Thereof
Advanced Corporate Finance – MBA 775A[pic 1]Department of Industrial and Management EngineeringINDIAN INSTITUTE OF TECHNOLOGY KANPURPatent Title A human-hand detection system, apparatus and a method thereof Indian patent application No.1678/DEL/2015Inventors nameDr. K. S. VenkateshMs. Garima JainMr. Vinod K Kurmi By: Shruti BansalRoll No: – 17125056ContentsInteraction with the inventor 3Details of the product 3Novelty of the invention. 4Applications of the technology 5Similar Technologies 5 Interaction with the inventor The inventor: Dr. K. S. VenkateshInteraction with the Dr. K. S. Venkatesh gave a lot of information regarding the patent. With the growing requirement for smart appliances SAMSUNG partnered with IIT Kanpur to deal with this requirement.
Details of the technology[1]The technology is a human detection system. The main objective of this technology is to provide a robust, convenient, reliable and fast detection of human hand irrespective of user postures and backgrounds. This technology also provides a mechanism for hand detection in real time irrespective of visible illumination levels, background 20 motion, foreground clutter, gesturing speeds and user distance. This technology will detect hand position/orientation under varying illumination conditions. The technology processes depth data for detecting a hand based on a 3D sensor data and thereby extracting the feature values corresponding to the hand detected.The technology uses a 3D sensor which is a Microsoft Kinect device, to capture the human hand with the user being in various poses and backgrounds. The Kinect sensor device includes a depth sensing technology, a built-in color camera, an infrared (IR) emitter, and a 5-microphone array, that enables the device to sense the location and movements of people as well as their voices. Further, the latest sensor, available in the markets provide up to three times higher depth fidelity with significant improvements in visualizing small objects and all objects more clearly. The Kinect sensor device is configured to project a pseudo-random pattern of the objects using IR rays, in order to calculate their 10 depth. In the Kinect depth map, each pixel value is mapped as a function of distance from the depth sensor. The depth data sent by the Kinect is of 16 bits, with the least significant 3 bits containing index of user detected at pixel and 1 bit for error. The remaining 12 bits carry the depth information. So, the actual depth at any pixel is related to the pixel value as follows: