An Open Access Journal
From: Vision-based vehicle detection and counting system using deep learning in highway scenes
Scenes | Video | Vehicle | Total number of vehicle objects | |||||
---|---|---|---|---|---|---|---|---|
frames | category | Our method | Full-image detection method | Actual number of vehicles | ||||
Remote area | Proximal area | Remote area | Proximal area | Remote area | Proximal area | |||
Scene 1 | 3,000 | Car | 6,128 | 8,430 | 493 | 6,616 | 6,849 | 8,550 |
Bus | 535 | 459 | 92 | 379 | 582 | 483 | ||
Truck | 5,311 | 5,320 | 840 | 4,703 | 5,792 | 5,471 | ||
Scene 2 | 3,000 | Car | 1,843 | 3,615 | 192 | 3,356 | 1,914 | 3,654 |
Bus | 194 | 364 | 82 | 295 | 207 | 382 | ||
Truck | 3,947 | 4,709 | 922 | 3,738 | 4,169 | 4,731 | ||
Scene 3 | 3,000 | Car | 1,774 | 2,336 | 224 | 2,188 | 1,834 | 2,352 |
Bus | 415 | 516 | 56 | 495 | 483 | 529 | ||
Truck | 3,678 | 3,490 | 731 | 2,662 | 3,726 | 3,507 |