• Home
  • About
    • Semih GULUM photo

      Semih GULUM

      Someone who aims to improve himself in deep learning and machine learning algorithms.

    • Learn More
    • Twitter
    • LinkedIn
    • Instagram
    • Github
  • Projects
  • Scrum Management
  • Certifications
  • Blog & Articles
  • CV
  • >

Mask, Cigarette, Smartphone and Sunglasses Detection. All in ONE!

28 May 2012

Reading time ~1 minute

Semih GULUM Mechatronic Engineer

Detection

List of libraries used

  • openCV
  • matplotlib
  • numpy
  • math
  • os
  • time
  • darknet
  • ctypes

Explanation the Project

Since I did not have a data set, I had to create it first. Therefore, I collected photos from social media platforms such as instagram, twitter and linkedin with the help of tags. As a result, I was able to create a dataset consisting of labeled 7462 data. This data set includes 1936 sunglasses, 3731 cigarettes, 450 smartphones and 2856 mask images.

Figure 1 - mask photos from every angle

But it is not that easy. We have a test! Lets see if it say mask or not.

Figure 2 - Simple Test For Mask Detection

Yeah! We pass the test :D Now let’s look at the others.

Figure 3 - Sunglasses Detect with Cool Matrix Sunglasses

Lol, if you can stop laughing, we can look at our next object, smartphones.

Figure 4 - Smartphone Detection

Our next object is a cigarette. However, I do not use smoking, this time I will ask my model to predict the images I found on the internet.

Figure 5 - Cigarette Detection

Thanks for reading my project!



Share Tweet +1