Visual dysfunction not only restricts the action ability of the vulnerable group,but also seriously affects their life,stud-y and work.In order to enable visually impaired people to read multimedia intuitively,naturally and friendly,a barrier-free multime-dia reader was designed based on image matching algorithm and feature point detection,and the performance of the system was veri-fied by image matching experiments and rotation invariance experiments.The results show that the combined feature point detection by image matching algorithm can quickly converge on the data,and can match 675 pairs per second,and the correct matching logarithm and matching accuracy are 179 pairs and 96.24%,respectively.In the case of large rotation,it has good rotation invariance,and vis-ually impaired people can quickly and accurately retrieve the target bibliography when they read barrier-free multimedia,which great-ly improves the reading experience.This study facilitates the study and life of the visually impaired and promotes the construction of a barrier-free and harmonious society.
关键词
图像匹配/多媒体阅读/无障碍/特征点检测
Key words
image matching/multimedia reading/barrier-free/feature point detection