A novel reading assistant system for blind, visually impaired people or children is developed in this paper. The user can choose the region to be read and then receive output via text to speech. We design a helpful system that users can choose the area (on the book) they want to listen when reading and the device can also speak out the entire article. Besides, users can read without distance perception by using this system. The step to construct this system is according to the following: first, this system is used depth images to detect the coordinates of the user's finger with Kinect, and the book is also input with Kinect. Next, we utilize text preprocessing to remove the border of the book and then take advantage of the common attribute of text height to eliminate images and tables. Beside, we also use a method to combine non-linear and linear compensation for correcting distortions of document images. Experimental results demonstrate that the proposed scheme has good performance and demonstrates robustness. It also provides an effective interactive platform.
關聯:
Journal of Applied Science and Engineering 21(3), p.475-484
