Fig 1.
Structure of the proposed vehicle detection system.
Fig 2.
Structure of the detection system.
Fig 3.
Structure of the feature extraction algorithm.
Fig 4.
Image cropping.
Fig 5.
Pre-processing (adjusting the contrast value).
Fig 6.
Result of a 3D plot in the region to be recognized.
Fig 7.
Process of composing a stacked DoG kernel.
Fig 8.
Comparison of the stacked DoG and patterns of the feature.
Fig 9.
Applying the kernel.
Fig 10.
Experimental results using the original AdaBoost classifier.
Fig 11.
Experimental results using the detection and tracking part of FCAAS.
Fig 12.
Experimental results using the proposed detection system.
Table 1.
Comparison of precision under some harsh environments.
Fig 13.
ROC’s of the three detection methods.
Fig 14.
Experimental results: Distinct-shaped vehicles.
Fig 15.
Experimental results: Vehicles loading large cargos.
Fig 16.
Experimental results: Vehicles without visible patterns.
Fig 17.
Experimental results: Vehicles in a tunnel.