System and method for alerting visually impaired users of nearby objects

A method for alerting a visually impaired user of collision, the method comprising:

• acquiring a first image and a second image, of a space in front of said user, by two spaced-apart imagers;
• deriving, from said first image and said second image, a three-dimensional reconstruction comprising a plurality of three-dimensional points;
• dividing said space in front of said user into a right collision frustum, a center collision frustum, and a left collision frustum, each having a three-dimensional location and orientation, relative to said user;
• further dividing each of said right collision frustum, said center collision frustum, and said left collision frustum into three collision frusta to provide a near right collision frustum, a near center collision frustum, a near left collision frustum, a medium right collision frustum, a medium center collision frustum, a medium left collision frustum, a far right collision frustum, a far center collision frustum and a far left collision frustum, said frusta being relative to said user;
• determining a presence of one or more objects proximate said user, based on counting a number of said plurality of three-dimensional points in each one of:
• said near right collision frustum, said near center collision frustum, said near left collision frustum, said medium right collision frustum, said medium center collision frustum, said medium left collision frustum, said far right collision frustum, said far center collision frustum and said far left collision frustum;
• when an object is determined to be present in adjacent collision frusta, a priority decision is made;
• and
• activating one of:
• a left, a center, and a right vibro-tactile device, in a particular way, based on said priority decision.

A system for assisting a visually impaired user comprising:

• a first imager and a second imager spaced apart producing a first image and a second image of a space in front of said user;
• a processing unit:
• receiving said first image and said second image, and deriving from said first image and said second image a three-dimensional reconstruction comprising a plurality of three-dimensional points,
• dividing said space in front of said user into a plurality of collision frusta each having a three-dimensional location and an orientation relative to said user for detecting one of:
• a presence of objects proximate a head of said user and an absence of a floor ahead of said user, and
• determining one of:
• said presence of objects proximate said head of said user and said absence of said floor ahead of said user based on counting a number of said plurality of three-dimensional points in each one of said plurality of collision frusta;
• and
• one or more controllers activating one or more vibro-tactile devices, in a particular way, in response to said determining by said processing unit;
• wherein said plurality of collision frusta comprises:
• a right collision frustum, a center collision frustum, and a left collision frustum, each representing a right, a center, and a left portion of said space in front of said user, respectively;
• and
• wherein said one or more vibro-tactile devices, comprise a right, a center, and a left vibro-tactile devices, and wherein said activating one or more vibro-tactile devices comprises:
• activating a right vibro-tactile device in response to said determining in said right collision frustum, activating a center vibro-tactile device in response to said determining in said center collision frustum, and activating a left vibro-tactile device in response to said determining in said left collision frustum.