Capture Photos From Glass Window Without Your Reflection In It

Those who love clicking snapshots of passing trees from a train's window or that of the beautiful river body seen from your plush hotel room, must have experienced the problem that the reflection of the indoors is captured in the picture and that spoils the entire effect. To tackle this issue, a team of researchers from MIT have developed a new algorithm that can distinguish reflection from the image and thus capture a clean photograph through a glass window. For years, professional shutterbugs have been using a workaround of wrapping their point-and-shoot camera's lenses with a dark cloth attached to the window with a tape. That kind of a setup isn't always feasible for on-the-go camera men.


By automatically removing reflections from digital photos, the new algorithm has made it easy to capture photos through glass double paned window. There are two nearly identical reflections, slightly offset from each other. The new technique involves and algorithm that divides images into 8-by-8 blocks of pixels; for each block, it calculates the correlation between each pixel and each of the others. The aggregate statistics for all these blocks in 50,000 training images could distinguish reflections from images shot through glass.

If such an algorithm is made highly robust, it can become part of everyday digital photogarphy as well as that of robotic vision. What are your thoughts about the new MIT algorithm? Share with us in comments below.

Source: #-Link-Snipped-#