On this page, some older Upview screen dumps are shown. More recent information on Upview can be found here.
Here is a file recorded in our own group (NICI.DAT), containing word-annotated cursive-connected words.
The image was produced with upview nici.dat
The next example shows character-annotated isolated handprint characters donated by the Hewlett Packard company, which funded part of the handwriting recognition research at NICI (HP1.DAT),
The colors of upview are stored in the control file $HOME/.upview Also note the use of ink width. Since in on-line recording the actual ink width is lost, it may be useful to add body to the characters. This can be done by clicking MoreInk or LessInk in upview.
The next example shows neatly written UK style cursive words (annotated at the word level) from Nottingham Trent University (NTU.DAT),
(Note: screen dumps were created by the GRAB function of xv)
Apart from the image of the on-line data, also the first derivative of the sequence of XY coordinates can be inspected. This gives clues about the tablet quality and the type of sampling (time or space equidistant), as this example file from the GO company shows (GO.DAT),
This image was produced by: upview -V go.dat
The last example comes from AT&T, i.e., Isabelle Guyon, who is leading the UNIPEN project (ATT.DAT), This file neatly implements the .HIERARCHY command of UNIPEN.
This image was produced by: upview att.dat
Back to the UNIPEN data and software page NICI