Vision-based Queue Monitoring System

Watch the demo video here: https://youtu.be/gUTV3HvWWVg.

Setting up

Clone the repo

git clone https://github.com/anhquan3012/dlw-queue-monitor.git
cd dlw-queue-monitor

Install mmdetection

Please install mmdetection library (mmdet) by following the instructions at https://mmdetection.readthedocs.io/en/latest/get_started.html#installation.

If you have problems with building pycocotools when trying to install mmdet, please install pycocotools first it by running this command.

conda install -c conda-forge pycocotools

After that you may install mmdet as per instructed.

git clone https://github.com/open-mmlab/mmdetection.git
cd mmdetection
pip install -v -e .

Install other dependencies

pip install -r requirements.txt

Install model configs and model weight

Our model use YOLOv3 with DarkNet53 Backbone with mixed precision training. Download the model configs and weights by running this command.

mim download mmdet --config  yolov3_d53_fp16_mstrain-608_273e_coco --dest .

You may use any model from mmdetection Model Zoo (https://mmdetection.readthedocs.io/en/latest/model_zoo.html#baselines). However, you will need to change the model file path inside the python files webcam_demo.py, or video_demo.py (which ever files you are going to use).

Running demo on webcam

You could run the demo on webcam by running webcam_demo.py.

python webcam_demo.py

Running demo on video

You could run the demo on a video by running video_demo.py (please lower the video frame rates and resolution before running).

python video_demo.py

Credit

Le Ky Nam

Pham Duc Nam Phuong

Pham Viet Hung

Phan Huu An

Ta Quynh Nga

Tran Anh Quan