Computer vision engineer Neel Joshi and his colleagues were going through some home photo collections and noticed that many pictures of faces were blurry.
"Precious moments were often lost due to blur induced by camera shake and poor lighting," says Neel.
Thus the team set off to make a software to correct these blur images. Their algorithm uses facial recognition methods, to find a sharp image with a similar pose and looks.
The pattern and shade of colour and light in the blurry face is then tweaked to match the model photo. Thus, correcting the pic and making it visible.
Now, no more of losing out on specials moments of life. This technology can also add colour and shine to black and white pics.