If you are using Windows, just double click on the downloaded jar file. If you are using Linux, right click on the file ocrevalUAtion.jar, select “open with”, and then the application called “Java 6 Runtime”. Then, A window like the one displayed below will open.
The window has two input areas (the white areas with dark text inside) where the user can drag and drop input files from the computer (open a navigator window and press the left mouse button over the input file icon while dragging it to the input area). Alternatively, the file can be selected using a standard selection window: click on “Or select the file” and a standard file selector window will open and allow one to navigate through the local file system.
Whenever a file is selected or dropped in the input area, the full name of the file appears. Please, note that:
- The first input is the ground truth file in any of the supported formats.
- The second is the OCR output (also in text, HTML or XML format).
Once all the input files has been uploaded, one can lick on the “Generate report” button. If either the required ground-truth or the OCR file is missing, then the associated area will be highlighted and the tool will wait till they are all selected. Once all the required input has been uploaded, the “Generate report” will open a window to select the name and location of the file which will contain the report. Since the report is generated in HTML format, it is highly recommendable that the the file name ends with the extension “.html”. Indeed, the tools suggests by default a file in the same directory as the OCR file and whose name ends with _report.html after the OCR file name.
Any attempt to overwrite an existing file will be flagged and asked for confirmation.
Finally, the program will open a browser showing the report. If, due to security restrictions, the system does not allow the program to open the browser, the output file can always be examined by clicking on it in a navigator window (using any browser such as Firefox, Chrome or MS-Explorer).