Real-Time Ghost Point Noise Reduction Algorithm for 2D Plane Array Scanning

To implement interactive user experiences like those found in immersive gaming, lightweight and fast player motion detection methods are being explored. One prominent approach involves using infrared to detect objects passing through a specific plane, known as plane detection. However, conventional 2D array scanning cannot simultaneously recognize more than two objects due to the ghost point phenomenon. The overlapping shadow regions of multiple objects create ambiguity, making it impossible for the system to identify the actual points. This study introduces a novel object point detection method for 2D array scanning. By devising a method comparing the axial sizes of object recognition data, real-time analysis of raw sensor data is conducted to remove noise and selectively identify actual touch points. The effectiveness of this method is validated through measurements of minimum detection size, simultaneous recognition of objects, and minimum distance between objects.