Detecting the Trajectory of Moving Object for Single-Pixel Imaging System

Abstract

Abstract. In order to get the trajectory of moving object using single-pixel imaging system, an algorithm is proposed. The same pseudorandom masks are employed to illuminate the different time scene. A time weighted sum of the background correction signals is employed to get the trajectory information using compressed sensing (CS) method. In ideal situation, we can obtain other parameters (e.g., speed, orientation) besides the trajectory. However, the reflective intensity of the object can be change due to the reflective angle change caused by the motion in some situations. This will mislead for achieving the speed, orientation parameters. In order to eliminate this effect, a division method is utilized. At last, the computer simulation results prove the effect validity of the proposed algorithm.