This project refers to https://github.com/Qidian213/deep_sort_yolov3. The camera and the video interface are offerd in this version. You can establish a folder model_data to put the pretrainded model and the the testing video.
If you want to test a video, please run the python demo.py.
If you want to use the camera ,please run the python demo.py --read_type=camera
Thanks for these projects, this work now is support tiny_yolo v3 but only for test, if you want to train you can either train a model in darknet or in the second following works. It also can tracks many objects in coco classes, so please note to modify the classes in yolo.py. besides, you also can use camera for testing.
https://github.com/nwojke/deep_sort
https://github.com/qqwweee/keras-yolo3
https://github.com/Qidian213/deep_sort_yolov3
- Download YOLOv3 or tiny_yolov3 weights from YOLO website.
- Convert the Darknet YOLO model to a Keras model.
- Run YOLO_DEEP_SORT
The following three steps, you can change accordingly:
please download the weights at first from yolo website or use your own weights.
python convert.py yolov3.cfg yolov3.weights model_data/yolo.h5
python demo.py
The code is compatible with Python 2.7 and 3. The following dependencies are needed to run the tracker:
NumPy
sklean
OpenCV
Additionally, feature generation requires TensorFlow-1.4.0.
file model_data/mars-small128.pb had convert to tensorflow-1.4.0
file model_data/yolo.h5 is to large to upload ,so you need convert it from Darknet Yolo model to a keras model by yourself
yolo.h5 model can download from https://drive.google.com/file/d/1uvXFacPnrSMw6ldWTyLLjGLETlEsUvcE/view?usp=sharing , use tensorflow1.4.0
speed : when only run yolo detection about 11-13 fps , after add deep_sort about 11.5 fps
test video : https://www.bilibili.com/video/av23500163/