Journal of Cloud Computing

Advances, Systems and Applications

Table 3 Pixel values of the rectangular detection boxes for each type of object

From: A deep reinforcement learning assisted task offloading and resource allocation approach towards self-driving object detection

Category	Pixel ranges	Thresholds
Pedestrian	width [30,50], length [100,200]	width 45, length 115
Car	width [180,220], length [180,400]	width 210, length 350
Truck	width [380,420], length [380,800]	width 410, length 720
Bicycle	width [20,200], length [180,220]	width 170, length 215

Back to article page