Occupany Tracker

Capacity trackering in real time in a bus. It is able to track the number of users with in a accumulated place.

How Yolov3 works

System Flow

Main Concepts

Screenshots

Dependencies

To run this project, you will need to add the following Dependencies to your file

absl-py numpy seaborn opencv-python scipy tensorflow matplotlib lxml tqdm

Running Tests

To run tests, run the following command

  python object_tracker_copy.py

Documentation

1. Pyenv : constantly changing python version in a devices
2. Pyenv github link : pyenv-win and version is necessary in this folder

Optimizations

For more optimize result , I considered yolov3 algorithm for training of data and considering different outcomes process for counting the passenger within an area

Roadmap

• python version: 3.7.9
• cuda version: 3.10v
• cudnn version: 3.7.6 for cuda version 3.10v
• GPU driver needed to get install for precision and accuracy

Installations

Save them to their weights folder

    https://pjreddie.com/media/files/yolov3.weights
    https://pjreddie.com/media/files/yolov3-tiny.weights

For checking if yolo's weights is loaded

yolov3

    python convert.py

yolov3-tiny

    python convert.py --weights ./weights/yolov3-tiny.weights --output ./weights/yolov3-tiny.tf

Checking GPU

import tensorflow as tf
print(tf.test.is_gpu_available(cuda_only =False,min_cuda_capability= None))