---------------------
CharImageManager Help
---------------------

CharImageManager is a tool to edit charcter cluster list obtained 
by the "MakeCharData" process of "BatchInfty". You can view the 
clustering results with the representative image of each cluster
grouped into the character codes of the recognition results.

1. Open

CharImaegManager can open "*.cim" files generated by InftyReader
Enterprise version.

********************************************************************
IMPORTANT REMARK:
When you open a "*.cim" file for the first time, please execute 
"Save All" in the "File" menu before editing the character clusters.
Otherwise, some clusters may be lost during the edition.
********************************************************************

2. Copy and move characters.

By clicking on character images, you can select or cancel selection
characters. If you click a character with "Shift" key pressed, you 
chan select all the characters between the clickt character and the 
character lastly selected. 

After the selection of wrongly classified characters, please select 
the correct character code on the left hand side code list (CODE VIEW), 
and then press the "MoveSelectedChar" button on the tool bar. Then, 
the selected characters are moved to the correct character code group.

3. Add new character code.

In case you cannot find the correct character code in the "CODE VIEW", 
please press the "AddCharacterCode" button on the tool bar. Selecting 
character cagegory and code from the list, a new character code can be 
added in the CODE VIEW list.

4. Make UserDic for InftyReader.

After the correction of all the errors of the clustering, please execute 
"Make UserDic" at the bottom of the "File" Menu, then "UserCharData.dat" 
is egnerated in the same folder as the cim file. 
If you copy the obtained "UserCharData.dat" to the same folder as the 
the corresponding "Ocrparameter.dat" used to make cim file, InftyReader 
and Infty.exe will run the recognition using the character features 
information included in the UserCharData.dat. Thus, you can correct the 
recognition results of all the characters in a cluster at once.

********************************************************************
IMPORTANT REMARK:
Please be careful that all the clusters are correctly copied into 
the corresponding character code group. In case there is a wrongly 
classified to different character code, all the character images in 
the target imaegs are wrongly recognized as it is asigned by the 
UserCharData.
*********************************************************************

Non Profit Organization 
Science Accessibility Net
2013.12.4

