ADVANCES IN MICRO-ELECTRONICS HAVE MADE COMPUTER-ASSISTED VISION POSSIBLE, BUT EXPENSIVE.

ADVANCES IN MICRO-ELECTRONICS HAVE MADE COMPUTER-ASSISTED VISION POSSIBLE, BUT EXPENSIVE. EVEN SIMPLE COMPUTER VISIONSYSTEMS COST IN EXCESS OF $5,000, WELL OUT OF THE ECONOMIC REACH OF THE BLIND, WHO ARE GENERALLY LOW PAID OR DEPENDENT ON GOVERNMENT SUPPORT. THIS PROJECT PROPOSES TO USE MAINLY EXISTING PRODUCTS WHOSE COSTS HAVE BEEN REDUCED BY HIGH VOLUME SALES. THE GOAL IS TO PRODUCE A DEVICE WITH SIGNIFICANT BENEFITS TO THE VISUALLY IMPAIRED THAT WILL SELLFOR LESS THAN $1,000. THE PROPOSED DEVICE, AN OPTICAL SCENEDESCRIBER (OSD), HAS BEEN PARTIALLY DEVELOPED. IT VIEWS A COARSE GRID REPRESENTATION OF A SCENE AND GIVES THE USER VERBAL AND TACTILE OUTPUT WITH REGARDS TO COLOR AND GEOMETRIC CONTENT AND DISTRIBUTION. THE USER WILL INTERACT WITH THE OSD TO REQUEST MORE DETAILED DESCRIPTION OF ANY PARTICULAR SCENE FEATURE. THE USER, NOT THE OSD, WILL PROVIDE THE CONTEXT AND COGNITIVE ABILITY TO UNDERSTAND THE SCENE. THE OBJECTIVES OF THE OSD WILL BE TO INCREASE USER COGNIZANCE AND MOBILITY AND, IF POSSIBLE, TO INCREASE USER PRODUCTIVITY AND EMPLOYMENT OPTIONS. THE RESEARCH FOR PHASE I WILL BE TO COMPLETE CONSTRUCTION OF THREE OSD'S AND CONDUCT FIELD STUDIES AT THE BRAILLE INSTITUTE. FEEDBACK FROM THE VISUALLY IMPAIRED EVALUATORS WILL BE USED TO IMPROVE THE DEVICE AND DEFINE WHICH ASPECTS NEED IMPROVEMENT OR ARE MOST USEFUL.