Detection and Tracking Dynamic Vehicles for Autonomous Driving Based on 2-D Point Scans

Optical Flow Based Detection and Tracking of Moving Objects for Autonomous Vehicles

“…• tracking performance relies on detection/classification • difficult to detect and track objects with unknown geometries Model-Free [10] [11] [12] g l [13] l [14] [20] [21] icp [22] icp [23] icp [24] icp [25] l [26] [27] g icp+l • tracking all scanned points • DATMO performance doesn't rely on geometric shape…”

“…This offers further insights to be incorporated into the downstream modules such as motion planning/prediction in the autonomous vehicle framework. The performance of the proposed technique compared to the ICP-based methods, such as the one in [21], [27], modelfree [14] and model-based indirect tracking methods [18], is evaluated in two steps. First, the compared DATMO techniques are evaluated on a synthetic dataset generated in MATLAB scenario designer, where various driving contexts are consid-ered.…”

“…There is a great number of research aimed to improve the object tracking task by developing enhanced classification/clustering steps (before tracking) using learning-based [12], [13] or geometric model estimation [1] algorithms, but all are still classified under the first category of DATMO methods. There are other studies in this category, such as [14], focused on reducing the computational complexity by applying the classification and subsequent tracking only on the moving point. In a detection-based approach the detection and tracking steps are independent and various sensor data could be used in the detection algorithm without changing the tracking part.…”

“…First and foremost we statistically compare the performance of the DATMO method with SOTA geometric model-free approach (GMFA) developed in [21] which has been proven to be more efficient than the geometric model-based tracking (MBT) method proposed in [37]. However, the proposed method is further compared against SOTA model free [14] and model-based [18] direct tracking methods as well. The GMFA algorithm is regenerated and evaluated, while the quantitative performance of the other methods are obtained from [14] and [18].…”

“…However, the proposed method is further compared against SOTA model free [14] and model-based [18] direct tracking methods as well. The GMFA algorithm is regenerated and evaluated, while the quantitative performance of the other methods are obtained from [14] and [18]. Therefore, the later experiments are consistent with those specified in these studies.…”