The implementation of an automatic training aid to analyse two dimensional images such as alphabetic characters and which could be used to feed back relevant information to the writer would offer a number of significant advantages in supporting the development of the coordination skills required in writing. In this context two principal issues have been tackled experimentally. The first deals with the characterisation and evaluation of the concept of 'quality' in character formation, and the convergence between perceived quality and the quantitative measures which can easily be generated in machine-based image analysis systems. The second issue concerns the question of how to exploit this type of information in providing writer feedback which can help to improve coordination in the execution of handwritten character formation.
展开▼
