Sign Language Recognition using Leap Motion Controller

摘要

Making deaf people use an alternative method instead of current voice input to 1CT equipment, we propose a sign language recognition method using Leap Motion Controller. As decisions using sign language, 16 kinds of decisions that focus on characteristic of hands and fingers are proposed. Sign language recognition algorithm is constructed using 16 kinds of decisions. The constructed flowchart is differing as order of decisions, recognition rate for all letters change from the difference accuracy rate of decisions. For sorting of decisions is enormous combination, genetic algorithm is applied to search for the optimal solution in the automatic construction of sign language recognition algorithm.