Skip to content
object detection using deep learning and multi-object tracking
Jupyter Notebook Shell
Branch: master
Clone or download
adipandas Merge pull request #4 from adipandas/developer
Bug fix for `RuntimeError: OrderedDict mutated during iteration bug` in Tracker class
Latest commit 0b15921 Nov 7, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
assets Tensorflow example with SSD-mobile-net Sep 14, 2019
caffemodel_dir Update with YOLO model in OpenCV Sep 8, 2019
tensorflow_model_dir Tensorflow example with SSD-mobile-net Sep 14, 2019
yolo_dir Update with YOLO model in OpenCV Sep 8, 2019
CODE_OF_CONDUCT.md Create CODE_OF_CONDUCT.md Nov 28, 2018
CONTRIBUTING.md Create CONTRIBUTING.md Nov 28, 2018
LICENSE Update LICENSE Oct 29, 2019
README.md
tracking-caffe-model.ipynb Bug fix Nov 7, 2019
tracking-tensorflow-ssd_mobilenet_v2_coco_2018_03_29.ipynb Bug fix Nov 7, 2019
tracking-yolo-model.ipynb Bug fix Nov 7, 2019

README.md

Multi-Object-Tracker

Object detection using deep learning and multi-object tracking

YOLO

Output Sample with YOLO

SSD

Output Sample with SSD

Install OpenCV

Pip install for OpenCV (version 3.4.3 or later) is available here and can be done with the following command:

pip install opencv-contrib-python

Run with YOLO

  1. Open the terminal
  2. Go to yolo_dir in this repository: cd ./yolo_dir
  3. Run: sudo chmod +x ./get_yolo.sh
  4. Run: ./get_yolo.sh

The model and the config files will be downloaded in ./yolo_dir. These will be used tracking-yolo-model.ipynb.

  • The video input can be specified in the cell named Initiate opencv video capture object in the notebook.
  • To make the source as the webcam, use video_src=0 else provide the path of the video file (example: video_src="/path/of/videofile.mp4").

Example video used in above demo: https://flic.kr/p/L6qyxj

Run with TensorFlow SSD model

  1. Open the terminal
  2. Go to the tensorflow_model_dir: cd ./tensorflow_model_dir
  3. Run: sudo chmod +x ./get_ssd_model.sh
  4. Run: ./get_ssd_model.sh

This will download model and config files in ./tensorflow_model_dir. These will be used tracking-tensorflow-ssd_mobilenet_v2_coco_2018_03_29.ipynb.

SSD-Mobilenet_v2_coco_2018_03_29 was used for this example. Other networks can be downloaded and ran: Go through tracking-tensorflow-ssd_mobilenet_v2_coco_2018_03_29.ipynb for more details.

  • The video input can be specified in the cell named Initiate opencv video capture object in the notebook.
  • To make the source as the webcam, use video_src=0 else provide the path of the video file (example: video_src="/path/of/videofile.mp4").

Video used in SSD-Mobilenet multi-object detection and tracking: https://flic.kr/p/26WeEWy

Run with Caffemodel

  • You have to use tracking-caffe-model.ipynb.
  • The model for use is provided in the folder named caffemodel_dir.
  • The video input can be specified in the cell named Initiate opencv video capture object in the notebook.
  • To make the source as the webcam, use video_src=0 else provide the path of the video file (example: video_src="/path/of/videofile.mp4").

References

The work here is based on the following literature available:

  1. http://elvera.nue.tu-berlin.de/files/1517Bochinski2017.pdf
  2. Pyimagesearch 1, 2
  3. correlationTracker
  4. Caffemodel zoo
  5. Caffemodel zoo GitHub
  6. YOLO v3

Use the caffemodel zoo from the reference [4,5] mentioned above to vary the CNN models and Play around with the codes.

Suggestion: If you are looking for speed go for SSD-mobilenet. If you are looking for accurracy and speed go with YOLO. The best way is to train and fine tune your models on your dataset. Although, Faster-RCNN gives more accurate object detections, you will have to compromise on the detection speed as it is slower as compared to YOLO.

Citation

If you use this repository in your work, please consider citing it with:

@misc{multiobjtracker_amd2018,
  author = {Deshpande, Aditya M.},
  title = {multi-object-tracker},
  year = {2018},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/adipandas/multi-object-tracker}},
}
You can’t perform that action at this time.