Current collected state Current
Observed since 20 Oct, 21:40 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
An easy-to-use Chinese license plate recognition system.
支持多种车牌的检测和识别(其中单层车牌识别效果较好):
单层车牌:

[[[[373, 282], [69, 284], [73, 188], [377, 185]], ['苏E05EV8', 0.9923506379127502]]]
[[[[393, 278], [318, 279], [318, 257], [393, 255]], ['VA30093', 0.7386096119880676]]]
[[[[[487, 366], [359, 372], [361, 331], [488, 324]], ['皖K66666', 0.9409016370773315]]]]
[[[[304, 500], [198, 498], [199, 467], [305, 468]], ['鲁QF02599', 0.995299220085144]]]
[[[[309, 219], [162, 223], [160, 181], [306, 177]], ['使198476', 0.9938704371452332]]]
[[[[957, 918], [772, 920], [771, 862], [956, 860]], ['陕A06725D', 0.9791222810745239]]]
双层车牌:
[[[[399, 298], [256, 301], [256, 232], [400, 230]], ['浙G66666', 0.8870148431461757]]]
[[[[398, 308], [228, 305], [227, 227], [398, 230]], ['陕A00087', 0.9578166644088313]]]
[[[[352, 234], [190, 244], [190, 171], [352, 161]], ['宁A66666', 0.9958433652812175]]]
快速安装
# 安装 AgentCLPR
$ pip install agentclpr
# 根据设备平台安装合适版本的 ONNXRuntime
# CPU 版本(推荐非 win10 系统,无 CUDA 支持的设备安装)
$ pip install onnxruntime
# GPU 版本(推荐有 CUDA 支持的设备安装)
$ pip install onnxruntime-gpu
# DirectML 版本(推荐 win10 系统的设备安装,可实现通用的显卡加速)
$ pip install onnxruntime-directml
# 更多版本的安装详情请参考 ONNXRuntime 官网
简单调用:
# 导入 CLPSystem 模块
from agentclpr import CLPSystem
# 初始化车牌识别模型
clp = CLPSystem()
# 使用模型对图像进行车牌识别
results = clp('test.jpg')
服务器部署:
启动 AgentCLPR Server 服务
$ agentclpr server
Python 调用
import cv2
import json
import base64
import requests
# 图片 Base64 编码
def cv2_to_base64(image):
data = cv2.imencode('.jpg', image)[1]
image_base64 = base64.b64encode(data.tobytes()).decode('UTF-8')
return image_base64
# 读取图片
image = cv2.imread('test.jpg')
image_base64 = cv2_to_base64(image)
# 构建请求数据
data = {
'image': image_base64
}
# 发送请求
url = "http://127.0.0.1:5000/ocr"
r = requests.post(url=url, data=json.dumps(data))
# 打印预测结果
print(r.json())
Download counts and popularity will appear here as the collector receives them.
| Version | Published | Direct dependencies | Open version |
|---|---|---|---|
| 1.1.0 | 2 Oct 2021 | 1 |
This describes our collected data; it is not a security assessment.
Observed since 20 Oct, 21:40 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
package:14179a7216b7ccc489ac3e61e8fa2786b91e25d5fcb53a2efa7d38683b6a69ee