Skip to content

Commit 8b032fb

Browse files
authored
cherry-pick of recently work (PaddlePaddle#7078)
* update empty features check (PaddlePaddle#6985) * rtsp push stream in pipeline (PaddlePaddle#7000) * add push stream * update push stream docs * update class name * update PP-Vehicle docs and en docs (PaddlePaddle#7042) * update PP-Vehicle docs and en docs * add reid model in v2.5 modelzoo * update more
1 parent ee563f0 commit 8b032fb

File tree

10 files changed

+215
-66
lines changed

10 files changed

+215
-66
lines changed

deploy/pipeline/README.md

Lines changed: 5 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -35,10 +35,10 @@
3535

3636
| ⭐ 功能 | 💟 方案优势 | 💡示例图 |
3737
| ---------- | ------------------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------- |
38-
| **车牌识别** | 超强性能:针对车辆密集、车牌大小不一问题进行优化,实现【待补充】 | <img title="" src="https://user-images.githubusercontent.com/48054808/185027987-6144cafd-0286-4c32-8425-7ab9515d1ec3.png" alt="" width="191"> |
39-
| **车辆属性分析** | 支持车型、颜色类别识别<br/><br/>兼容多种数据格式:支持图片、视频、在线视频流输入<br/><br/>高性能:融合开源数据集与企业真实数据进行训练,实现【待补充】<br/><br/> | <img title="" src="https://user-images.githubusercontent.com/48054808/185044490-00edd930-1885-4e79-b3d4-3a39a77dea93.gif" alt="" width="207"> |
40-
| **违章检测** | 易用性高:一行命令即可实现违停检测<br/><br/>鲁棒性强:对光照、视角、背景环境无限制 | <img title="" src="https://user-images.githubusercontent.com/48054808/185028419-58ae0af8-a035-42e7-9583-25f5e4ce0169.png" alt="" width="209"> |
41-
| **车流量计数** | 一键运行:单个参数即可开启车流量计数与轨迹记录功能 | <img title="" src="https://user-images.githubusercontent.com/48054808/185028798-9e07379f-7486-4266-9d27-3aec943593e0.gif" alt="" width="200"> |
38+
| **车牌识别** | 支持传统车牌和新能源绿色车牌 <br/><br/> 车牌识别采用长间隔采样识别与多次结果统计投票方式,算力消耗少,识别精度高,结果稳定性好。 检测模型 hmean: 0.979; 识别模型 acc: 0.773 | <img title="" src="https://user-images.githubusercontent.com/48054808/185027987-6144cafd-0286-4c32-8425-7ab9515d1ec3.png" alt="" width="191"> |
39+
| **车辆属性分析** | 支持多种车型、颜色类别识别 <br/><br/> 使用更强力的Backbone模型PP-HGNet、PP-LCNet,精度高、速度快。识别精度: 90.81 | <img title="" src="https://user-images.githubusercontent.com/48054808/185044490-00edd930-1885-4e79-b3d4-3a39a77dea93.gif" alt="" width="207"> |
40+
| **违章检测** | 简单易用:一行命令即可实现违停检测,自定义设置区域 <br/><br/> 检测、跟踪效果好,可实现违停车辆车牌识别 | <img title="" src="https://user-images.githubusercontent.com/48054808/185028419-58ae0af8-a035-42e7-9583-25f5e4ce0169.png" alt="" width="209"> |
41+
| **车流量计数** | 简单易用:一行命令即可开启功能,自定义出入位置 <br/><br/> 可提供目标跟踪轨迹显示,统计准确度高 | <img title="" src="https://user-images.githubusercontent.com/48054808/185028798-9e07379f-7486-4266-9d27-3aec943593e0.gif" alt="" width="200"> |
4242

4343
## 🗳 模型库
4444

@@ -53,6 +53,7 @@
5353
| 行人检测(轻量级) | 16.2ms | [多目标跟踪](https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_s_36e_pipeline.zip) | 27M |
5454
| 行人跟踪(高精度) | 31.8ms | [多目标跟踪](https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_l_36e_pipeline.zip) | 182M |
5555
| 行人跟踪(轻量级) | 21.0ms | [多目标跟踪](https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_s_36e_pipeline.zip) | 27M |
56+
| 跨镜跟踪(REID) | 单人1.5ms | [REID](https://bj.bcebos.com/v1/paddledet/models/pipeline/reid_model.zip) | REID:92M |
5657
| 属性识别(高精度) | 单人8.5ms | [目标检测](https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_l_36e_pipeline.zip)<br> [属性识别](https://bj.bcebos.com/v1/paddledet/models/pipeline/strongbaseline_r50_30e_pa100k.zip) | 目标检测:182M<br>属性识别:86M |
5758
| 属性识别(轻量级) | 单人7.1ms | [目标检测](https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_l_36e_pipeline.zip)<br> [属性识别](https://bj.bcebos.com/v1/paddledet/models/pipeline/strongbaseline_r50_30e_pa100k.zip) | 目标检测:182M<br>属性识别:86M |
5859
| 摔倒识别 | 单人10ms | [多目标跟踪](https://bj.bcebos.com/v1/paddledet/models/pipeline/mot_ppyoloe_l_36e_pipeline.zip) <br> [关键点检测](https://bj.bcebos.com/v1/paddledet/models/pipeline/dark_hrnet_w32_256x192.zip) <br> [基于关键点行为识别](https://bj.bcebos.com/v1/paddledet/models/pipeline/STGCN.zip) | 多目标跟踪:182M<br>关键点检测:101M<br>基于关键点行为识别:21.8M |

0 commit comments

Comments
 (0)