Robust and low-cost proximity sensor for line detection robot using goertzel algorithm

Abstract

The usage of proximity sensor for marker detections such as color line detection is the easiest technique in robotic competition. Most of the contestants use simple photo-reflective circuit as proximity sensor with constant light source and voltage-based threshold to deduce the presence of marker line. By using this technique, noises such as various changes in indoor lighting exposure and camera flash may cause negative effect against system performance. To overcome this situation, a proximity sensor model is presented utilizing photo-reflective circuit with modulated light source. Goertzel algorithm is then used to calculate the power spectrum value of the received signal from photo-reflective circuit. The presence of marker line is deduced by simple thresholding technique based on the previous calibrated power spectrum value. Using this model, marker line detection method based on proximity sensor becomes more robust against noises and disturbances.