Detecting Ghost Targets Using Multilayer Perceptron in Multiple-Target Tracking

Abstract

This paper deals with a method for removing a ghost target that is not a real object from the output of a multiple object-tracking algorithm. This method uses an artificial neural network (multilayer perceptron) and introduces a structure, learning, verification, and evaluation method for the artificial neural network. The implemented system was tested at an intersection in a city center. Results from a 28-min measurement were 88% accurate when the multilayer perceptron for ghost target classification successfully detected the ghost targets, and 6.7% inaccurate when ghost targets were mistaken for actual targets. This method is expected to contribute to the advancement of intelligent transportation systems if the weaknesses revealed during the evaluation of the system are complemented and refined.

Super resolution passive radars based on 802.11ax Wi‐Fi signals for human movement detection

“…We remind that the ghost targets are high‐order reflections caused by the real targets in the environment, which is common when indoor scenarios are considered. However, the ghost targets can be eliminated by using a tracking algorithm [28] or a multi‐static radar configuration [29].…”

An Elaborated Signal Model for Simultaneous Range and Vector Velocity Estimation in FMCW Radar

“…For correct use of the method of two-dimensional Fourier transform [ 19 , 20 , 21 ] and an unambiguous estimate of the velocity vector of a radar target, it is necessary that the phase change Φ V ( kT c ), calculated on the local time interval [0 ≤ t L ≤ T c ], does not exceed the value π. This requirement determines that the point θ belongs to the three-dimensional subspace Θ ( 3) 2 of the measured parameters { V R , V TR , R 2 } in the relation Equation (34).…”

“…The creation of a segmented structure of the sounding radio signal can significantly reduce the likelihood of false targets. Phantom targets are eliminated in the process of monitoring the radar situation using tools of intelligent data analysis technologies (Data Mining), such as fuzzy logic, neural networks and Kalman filtering [ 18 , 19 ]. Solving the problem of false radar targets inevitably leads to an increase in the duration of observation and data analysis, which limits the application area of LFMCW radars that use the one-dimensional (1D) Fourier transform algorithm.…”

Extending Reliability of mmWave Radar Tracking and Detection via Fusion With Camera

“…For the first case, which is the multipath effect in 77 GHz automotive radar, a typical symptom is the presence of ghost targets, that are produced with random phase but similar shape, Doppler and amplitude compared to the original target. A method to reduce when applied to tracking is to use a multilayer perceptron [30]. In this case, the ghost is largely alleviated and then the tracking becoming more reliable.…”