AIAS/vehicle_sdk at d96ca7ee104b00c2a7cbed28dfd6e992e1f3f1f0 - AIAS

test/AIAS

mirror of https://gitee.com/mymagicpower/AIAS.git synced 2024-11-30 03:08:24 +08:00

History

Calvin 9c5708095c update english comments		2023-03-21 20:46:31 +08:00
..
doc/img	no message	2021-10-03 22:25:35 +08:00
models	load model from local disk instead of http url	2023-03-20 13:56:52 +08:00
src	update english comments	2023-03-21 20:46:31 +08:00
pom.xml	upgrade to 0.17.0	2022-05-28 14:48:33 +08:00
README_cn.md	initial version - writing english doc.	2023-03-20 17:07:14 +08:00
README.md	update readme.	2023-03-20 23:02:51 +08:00
vehicle_sdk.iml	no message	2021-10-03 22:25:35 +08:00
vehicle-sdk.iml	upgrade to 0.17.0	2022-05-28 14:48:33 +08:00

README.md

Download the model and place it in the models folder

Link: https://github.com/mymagicpower/AIAS/releases/download/apps/vehicle.zip

Vehicle Detection SDK

Vehicle detection is a very important and challenging task in urban traffic monitoring, and its difficulty lies in accurately locating and classifying relatively small vehicles in complex scenes. Recognize vehicle types such as car, truck, bus, motorbike, and tricycle.

SDK functions

Vehicle detection, giving detection boxes and confidence
Supported categories: -car -truck -bus -motorbike -tricycle -carplate

Example-VehicleDetectExample

Test image

After running successfully, the command line should see the following information:

[INFO ] - [
	class: "car", probability: 0.98734, bounds: [x=0.210, y=0.420, width=0.225, height=0.218]
	class: "car", probability: 0.93550, bounds: [x=0.377, y=0.432, width=0.150, height=0.120]
	class: "car", probability: 0.88870, bounds: [x=0.167, y=0.411, width=0.127, height=0.178]
	class: "car", probability: 0.85094, bounds: [x=0.479, y=0.351, width=0.504, height=0.476]
	class: "carplate", probability: 0.83096, bounds: [x=0.321, y=0.502, width=0.046, height=0.019]
]

Open source algorithm

1. Open source algorithms used by the SDK

PaddleDetection