update

2024-10-15 22:25:25 +08:00 · 2024-10-15 22:25:25 +08:00 · 1566716165
parent 0e6d01bc77
commit 1566716165
17 changed files with 6 additions and 2 deletions
--- a/README.md
+++ b/README.md
@ -1,3 +1,7 @@
+---
+typora-copy-images-to: ./images
+---
+
 ## Yolov5s+SORT(卡尔曼+匈牙利)

 利用目标跟踪对检测框进行预测的技术，对目标检测任务进行加速，在yolov5检测的帧之间穿插使用卡尔曼滤波进行目标检测框的预测，减少对模型的调用，实现目标检测帧率提升的目的。
@ -15,11 +19,11 @@

 main.py中

-![image-20241014170146823](C:\Users\FengHua\AppData\Roaming\Typora\typora-user-images\image-20241014170146823.png)
+![屏幕截图 2024-10-14 170104](./images/屏幕截图 2024-10-14 170104-1729002234953-3.png)

 使用到的模型以及检测视频的路径。

-![image-20241014170213505](C:\Users\FengHua\AppData\Roaming\Typora\typora-user-images\image-20241014170213505.png)
+![屏幕截图 2024-10-14 170210](./images/屏幕截图 2024-10-14 170210-1729002243204-5.png)

 模型的调用频率设置，设置为3表示帧率为3的倍数时进行模型的目标检测，其他时候使用预测的目标框。

--- a/images/1.mp4
+++ b/images/1.mp4
--- a/images/2.mp4
+++ b/images/2.mp4
--- a/images/bus.jpg
+++ b/images/bus.jpg
--- a/images/detect.mp4
+++ b/images/detect.mp4
--- a/images/out1.mp4
+++ b/images/out1.mp4
--- a/images/out2.mp4
+++ b/images/out2.mp4
--- a/images/res.mp4
+++ b/images/res.mp4
--- a/images/res1.mp4
+++ b/images/res1.mp4
--- a/images/test.mp4
+++ b/images/test.mp4
--- a/images/zidane.jpg
+++ b/images/zidane.jpg
--- a/170104-1729002234953-3.png
+++ b/170104-1729002234953-3.png
--- a/images/屏幕截图
+++ b/images/屏幕截图
--- a/170210-1729002243204-5.png
+++ b/170210-1729002243204-5.png
--- a/images/屏幕截图
+++ b/images/屏幕截图
--- a/models/best.onnx
+++ b/models/best.onnx
--- a/models/put_model_here
+++ b/models/put_model_here