mirror of
https://gitee.com/mymagicpower/AIAS.git
synced 2024-12-02 04:08:21 +08:00
1.9 KiB
1.9 KiB
Download the model and place it in the /models directory
Pedestrian Detection SDK
Pedestrian detection is the use of computer vision technology to determine whether there are pedestrians in the image and to provide accurate positioning, generally represented by a rectangular box. Pedestrian detection technology has strong practical value. It can be combined with pedestrian tracking, pedestrian re-identification, and other technologies, and applied to autonomous driving systems for cars, intelligent video surveillance, human behavior analysis, passenger flow statistics system, intelligent transportation and other fields.
SDK Functionality
- Pedestrian detection, providing detection boxes and confidence
Running Example- PedestrianDetectExample
- Test image
After successful operation, the command line should see the following information:
[INFO ] - [
class: "pedestrian", probability: 0.97251, bounds: [x=0.284, y=0.451, width=0.101, height=0.394]
class: "pedestrian", probability: 0.97015, bounds: [x=0.418, y=0.448, width=0.082, height=0.377]
class: "pedestrian", probability: 0.96476, bounds: [x=0.568, y=0.423, width=0.105, height=0.411]
class: "pedestrian", probability: 0.95523, bounds: [x=0.811, y=0.401, width=0.104, height=0.436]
class: "pedestrian", probability: 0.93908, bounds: [x=0.680, y=0.433, width=0.074, height=0.352]
]