TensorFlow Machine Learning Image Recognition Python API Tutorial

TensorFlow is an open-source software library for dataflow programming across a range of tasks. It is a symbolic math library and is also used for machine learning applications such as neural networks.

It is the fastest and simplest way to do image recognition using CPU with inception-v3 on your computer without any GPU because it is just a Python API.

In this tutorial, I will let you know how to perform TensorFlow Machine Learning Image Recognition to classify your image. It is just very easy to recognize the image by using a pre-trained model.

Download Docker

Go to the docker official website and download docker community. After installing the docker, open it on your computer.

//www.docker.com/community-edition#/download

Install and run Tensorflow via Docker

In your terminal enter the command below to install and start tensorflow with CPU.

docker run -it -p 8888:8888 tensorflow/tensorflow

Cloning the TensorFlow model’s repo from GitHub

Open your terminal and insert the following commands line.

git clone github.com/tensorflow/models.git
cd models/

Run the classify_image.py file

In your terminal run command to run classify_image.py, this will download a pre-trained model online which will help you in recognizing your image later.

cd tutorials/image/imagenet/
python classify_image.py

Download image from google

Go to google and download image from this link //www.google.com/search?q=toy+dog&safe=active&tbm=isch&tbs=rimg:CQUhWmMfQ3IBIjha3hCRfs6gkCx1C0PyJxdBRRjs2pkeu3l-z7qRb-XJ-aCnOIh4tTZ5lYN4lv1X2zljgVC6-mkyLSoSCVreEJF-zqCQETYHpcIrCIRzKhIJLHULQ_1InF0ERQzrqU62lQNsqEglFGOzamR67eRHxSs9SXdXylioSCX7PupFv5cn5ETnnN9Lz5Ne4KhIJoKc4iHi1NnkRpItxKHILo-QqEgmVg3iW_1VfbORFFepFlQMMDxioSCWOBULr6aTItEZYZHWYy1LhG&tbo=u&sa=X&ved=2ahUKEwistPyn7dLcAhUYT30KHRy5CpYQ9C96BAgBEBs&biw=1440&bih=726&dpr=1#imgrc=BSFaYx9DcgEzAM:

Perform Image Recognition

Enter the following command line to perform image recognition function via python.

python classify_image.py --image_file Toy-dog-bread-playing.jpg

TensorFlow Machine Learning Image Recognition Python API Results

You can see the result in your terminal after you execute the image recognition command.

Pomeranian (score = 0.98215)

keeshond (score = 0.00336)

chow, chow chow (score = 0.00095)

toy poodle (score = 0.00026)

papillon (score = 0.00024)

Yeah, now you have done the image recognition in your laptop within few seconds to get the results.

(Visited 621 times, 1 visits today)
Advertisements

Yong Loon Ng

Ng Yong Loon, better known as Kristofer is a software engineer and computer scientist who doubles up as an entrepreneur.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *