- Massachusetts Institute of Technology, Microsoft and Abobe have together developed a method of reconstructing sound from the video of an object.
- When sound hits an object, it causes the object to vibrate, this vibration creates a very subtle signal that is invisible to the naked eye but can be detected using a high-speed video.
- The technique is similar to the laser microphones used by spies to eavesdrop on conversations by measuring minute vibrations in reflective surface.
- But unlike the sophisticated instruments used by laser microphones, this technique uses simple objects such as a crisp packet, a glass of water or potted plant as a microphone.
- Applications lie in surveillance and law enforcement, and in investigating the internal make up of objects by using their acoustic properties.
- The algorithm that extracts the information requires a very high-speed video where the frequency of audible sound ranges from 20 to 20,000Hz.
- Only a camera with a frame rate higher than this frequency can be used to extract the sound. Conventional cameras with a frame rate of 60 frames per second can only extract extremely low-frequency sounds.
- The researchers have developed another technique, ‘rolling shutter’ which don’t save the whole image in one go, but line-by-line so as to extract more information.
- This allows us to extract the sound played to a video recording of crisp packet from a simple digital SLR camera, shooting video of a crisp packet at 60fps.
- Read at: http://www.thehindu.com/todays-paper/tp-national/how-a-chips-packet-can-be-used-to-eavesdrop/article6293343.ece