https://github.com/onnesok/machine-vision

tensorflow object detection codes. This project can be implemented in many automation and robotics projects.
https://github.com/onnesok/machine-vision

machine-vision opencv tensorflow tensorflow2

Last synced: about 2 months ago
JSON representation

tensorflow object detection codes. This project can be implemented in many automation and robotics projects.

• Host: GitHub
• URL: https://github.com/onnesok/machine-vision
• Owner: Onnesok
• Created: 2024-05-27T10:19:49.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-06-16T07:22:37.000Z (8 months ago)
• Last Synced: 2024-11-09T16:12:16.472Z (3 months ago)
• Topics: machine-vision, opencv, tensorflow, tensorflow2
• Language: Python
• Homepage:
• Size: 1.07 MB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
##  Tensorflow

![copy](https://www.tensorflow.org/images/tf_logo_horizontal.png)

 

 This repository contains basic codes of tensorflow object detection. It's maintained for a baseline code start.

[![Compatibility](https://img.shields.io/badge/python-3.12-brightgreen.svg)](https://www.python.org/)

[![Modified](https://img.shields.io/badge/Coverage-working-red)](machine-vision)

[![Hits](https://hits.sh/github.com/Onnesok/machine-vision.svg)](https://hits.sh/github.com/Onnesok/machine-vision/)

## Installation

First create a virtual environment named tf_vision to avoid package conflicts. use ``cmd or bash`` to do it.

```bash

python3 -m venv tf_vision

```

Now source or activate environment.

#### For linux

```bash

source ./tf_vision/bin/activate

```

#### For windows cmd

```bash

tf_vision\Scripts\activate

```

## Install tensorflow and opencv

For latest version......

```bash

pip install tensorflow

```

```bash

pip install opencv-python

```

wait.... select interpreter of vs code to your selected virtual environment from the bottom right corner.

But it will run using cpu resources. So, lets run using ``gpu`` so that it can use more cores.

But, native windows does not support gpu on tensorflow version greater than ``2.10.0``..........Also, package seems to be removed right now for ``tensorflow-gpu==2.10.0`` using ``pip``. 

For compatiblity see here ==> https://www.tensorflow.org/install/source_windows

So, lets do it using ``conda``

Create virtual env using conda first and activate.

```bash

conda create -n conda_env python==3.9

conda activate conda_env

```

Now, select interpreter of vs code to your selected virtual environment from the bottom right corner.

Ok, now install cudatoolkit and cudnn for gpu support.

```bash

conda install cudatoolkit=11.2 cudnn=8.1 -c=conda-forge

```

Now install tensorflow-gpu version 2.10.0 and opencv-python

```bash

pip install tensorflow-gpu==2.10.0

pip install opencv-python

```

Done.