ENABLING SPEECH SYNTHESIS

Thanks to the speech synthesis, you can make the IRISPen speak out the recognized data for you*. Just scan some data and the computer reads it out for you!

(*) The speech synthesis feature is available in English.

Practically, you will no longer need to look up to the screen to verify your recognition results. Think how useful this will be when encoding new data, entering data into your bookkeeping, etc. The IRISPen will also be a valued tool when learning a foreign language: the computer will pronounce all new words for you, etc.

At any moment in time, you can also have the software repeat the pronounced lines. The IRISPen will always repeat the last utterance.

Tip:

Speech synthesis applies to all legible data

Speech synthesis not only applies to text but to all legible data such as handprinting, numeric data, and barcodes. You can thus activate the speech to get an auditive feedback of any type of data.

Tip:

Do not enable the speech feature unless you really intend to use it. The recognition actually slows down as you wait for the recognized data to be read aloud.

Tip:

If you intend to read whole sentences, do not forget to select the scanning mode Multiple lines. This will ensure a natural prosody and a good intonation.

To activate the speech synthesis:

• Make sure that English has been selected as recognition language. See: Basic parameters, Selecting the language.
• If you intend to read whole sentences, select the scanning mode Multiple lines. This will ensure a natural prosody, a good intonation, and the insertion of pauses. See: Additional options, Multiple lines.

• Click on Enable Speech in the IRISPen Palette or select Speech from the Options menu. As soon as data is being scanned, it is spoken out by your computer.

If the beep sound was enabled, it is automatically disabled and replaced by the speech.

• To repeat the speech:

Select the command Repeat Speech from the Options menu. The IRISPen will directly repeat the last pronounced utterance for you.

ADJUSTING THE IMAGE

The image adjustment features enable to optimize the image generated with the pen for the recognition process. These options will be particularly useful if you scan very bad quality documents, documents with colored background, or low-contrast documents where the background virtually has the same color as the text.

After each adjustment, the scanned image is directly adapted and the recognition re-executed.

Tip:

Remember that yellow is the IRISPen's invisible color. Do not try to make the IRISPen pick up any data printed in yellow.

To adjust the scanned image:

• Click on Adjust the Image in the IRISPen Palette or select

Adjust Image from the Options menu.

Adjust the settings as follows. The resulting image is directly dislayed in the Scan window.

Smoothen image

The Smoothen image option renders the images more homogeneous by "flattening", smoothening out relative differences in intensity. As a result, a sharper contrast is created between the text and the background.

The option is activated by default. It is recommended to leave it activated at all times.

Brightness

Brightness means the black-and-white threshold of the image.

Automatic brightness is set by default. You should leave this default setting for all normal documents.

Manual

When dealing with bad quality documents, changing manually the brightness level may level the quality of the black&white image and so improve the text recognition.

To set the brightness manually, click Manual, then slide the indicator to the left (lighten) or to the right (darken). When you darken the image, more pixels become black in the black-and-white image; when you lighten the image, less pixels become black.

Despeckle

The option Despeckle makes it possible to remove small spots from the images. By default, the option is turned off.

To activate the despeckling, slide the indicator to the right. You can remove dots from 1 to 20 pixels.

! Caution:

Be careful not to erase spots that are too big, otherwise you may start erasing the dots on the letter "i", portions of dot matrix letters, etc.