Binarization of Call Number Images For Helping Elderly Retired Volunteer to Manage Books in Library
Chun-Ming Tsai, Tawei David Shou, Jun-Wei Hsieh, Mu-Tsai Chang · 2018
All libraries use call numbers (CNs) on the spine to organize books. However, the CNs are small and often cannot be seen clearly by people with impaired sight, especially senior citizens. This makes it hard for seniors to find books they want. Moreover, retired seniors often work in libraries as volunteers to return books to the shelves. However, seniors, reduced vision often means the volunteers readers cannot see the CNs clear enough to replace the book correctly and senior readers cannot rmd the book they want. In order to help senior citizens both as readers and retired senior volunteers, a method to threshold characters in CN images is proposed. The proposed method includes color image transformed into gray level image, border histogram computation, histogram smoothing, major peaks finding, threshold value determination, and gray level image thresholded into binary image. From the binarization and OCR results it is demonstrated that the proposed method can threshold the characters in CN images with two backgrounds and non-uniform background.