The objective of our project is to design a book scanner at an accessible price. There are applications for which it is not worth spending a fortune getting a very high tech device to scan books, but a cheaper one could be considered. This mainly concerns situations in which book scanning is not essential but would make life easier. People have been seen spending hours copying someone’s notebook page by page with a traditional scanner, or trying to take an acceptable picture of their mum’s old recipes. Even unconditional book lovers who do not want to give up the feeling of holding a real book in their hands have faced the awful dilemma of which book they should bring on holidays because they can not carry them all. For all these situations and many others, a cheap even do-it-yourself and easy to use book scanning device would bring a good solution, allowing the user to always carry his favourite books and recipes with him on his portable devices. Besides, the scanner could also be used to numerize important book collections without having to invest too much.
We first considered the option of filming the book while turning its pages in order to make it easier for the user and more fluid. Though the first tests had shown that it was possible to get readable images using this technique further investigations showed that the loss of quality between a picture and filming with the same device was too important and it was decided to use pictures. The other advantage of it is that it makes the software part of the project much simpler and therefore more fluid, letting less possibility to eventual bugs.
Figure 1: Images of a book taken with the same camera taking a picture (right) and from a video (left).
The most expensive part of the scanner being the camera we thought that a good way to reduce the cost would be to simply use the camera mostly everyone nowadays has: the camera of a smartphone. Recent smartphones have quite good resolution, the pictures in figure 1 were actually taken using a smartphone camera. The scanner would then consist of an adjustable support for the smartphone, a support for the book and a pedal connected to the smartphone which the user would use to take the pictures of each page. This configuration lets the user’s hands free so he can use them to turn the pages and hold the book flat.
For the book support our first idea was to use a “V” shape but after some tests we figured out it was easier and more efficient to simply hold the book completely opened. We are therefore using a flat support with a gap in the middle for the bookbinding to keep the book from moving. This position of the book allows to take both pages at once with a single camera while keeping the central part of the book readable which was not really achieved with the “V” shape.
The next steps of the project will be to develop a first prototype of the scanner and an app for the smartphone which would work with the pedal, take the pictures, reshape them and store them in the format we want.
Gabrielle Tremblay Otis