A novel infrared touch sensing using K-nearest neighbor algorithm

Production and usage of touch sensing has attracted much attention, nevertheless, the current touch screen systems highly suffer from the dependency on the platform material. To overcome this drawback, here, a novel method is proposed by considering the infrared transmitters and receivers sensors instead of the piezoelectric sensors. In this way, the transformed platform works based on employing K-nearest neighbor (KNN) in conjunction with the spatial-temporal windowing to increase the sensitivity and accuracy of the detecting the finger place on the screen. To verify the proposed approach, the necessary hardware was set up. To test the hardware, the platform is locally segmented (e.g. like a grid) into a 7*8 table and 50 persons with different finger size were participated. Experimental results show the superior efficiency of the proposed method compare to the conventional methods.