Image Focus Recognition and Classification Neural Network

Photographs are used to capture and record detailed still moments of different scenarios, environments or subjects for a wide variety of purposes. For example, taking a high quality picture of a large celestial body, like the moon in order to see and study its craters for scientific research, to even capturing microscopic images of tiny micro-organisms to study how they interact with their surrounding world, to even just capturing a special moment in a person's life, like a wedding, or a vacation. Regardless of the application of the images, it is important to capture as much detail as possible, but this detail can be lost when the image is out-of-focus; Not only that, it could also render the image unusable.

Machine learning has been used in the past not only for binary classification of in- and out-of-focus images, but also the classification of different kinds of blur. The goal for this project was to classify images that are in-focus and out-of-focus using Neural Networks. This could have useful applications in a camera focusing system, aiding in either manually focusing the lens, or having it connected to a motorised focusing system and adjusting focus automatically.

I've also made a demonstartion of me using the Neural Network in this youtube video!