This document outlines the basic steps for using the OCR (Optical Character Recognition) function in UnaGPT. This feature allows you to convert images of handwritten text into editable, digital text, which can then be used to submit student answers or input new questions directly into the system.

OCR input text

Several subjects are available for OCR scanning. (The available OCR buttons depend on your system’s configuration.)

1. English
2. Chinese
3. Math
4. Code

Using the OCR Function

For the input field, there should be an “OCR (…)” button which sends the recognized text to that field. Click this button to open a pop-up window where you can scan images containing handwritten text.

Importing Images

There are two ways to import images:

1. Red Area (Image Files): Click “Choose Files” to select image files from your device.
2. Green Area (Clipboard): Copy an image from another application (e.g., a screenshot) and it will be automatically pasted into the clipboard. The OCR window will detect the image data.

Click the “X” button anytime to cancel the OCR function and close the pop-up window.

Editing Images for Optimal OCR

You can edit uploaded images to improve OCR accuracy.

• Red Area (Image Order): Indicates the order in which images will be processed and their text concatenated. Drag and drop images to reorder them.
• Blue Area (Image Cropping and Rotation): Select a specific area of the image to focus on and adjust the image angle if necessary. This is useful for removing irrelevant parts of the image or correcting skewed images. By default, the entire image is used.
• Green Area (Zoom and Pan): On non-touch devices, use the zoom and pan buttons to adjust the image view for better precision.
• Yellow Area (Perform OCR): Click “Perform OCR” to start the text recognition process. The processing time may vary depending on network conditions and the image complexity.
• Black Area (Add Images): Click “Add Images” to upload additional files. The text extracted from these images will be added to the existing output.

You can safely minimize the pop-up window during OCR processing to continue working on other questions or tasks. This will not close the OCR window or discard your progress.

Reviewing and Editing the OCR Output

Once the OCR process is complete:

1. Review the recognized text in the text box on the right side of the window.
2. Edit the text to correct any errors.
3. Click “Copy & Append” to add the corrected text to the answer box in UnaGPT.

OCR Math Mode – Special Functions

The Math mode offers specialized features for working with mathematical expressions:

You can click the buttons with math formulas to edit the text. Alternatively, you can open the editor by clicking the “Insert Math Content” button to add a new formula.

• Orange Area: Click on the formula to edit mathematical expressions.
• Yellow Area: Click “Insert Math Content” to add missing mathematical expressions.
• Use the “Undo” button to revert the most recent action. (Note: Only the last action can be undone.)

Math Editor Interface

• Blue Area: Click to open or hide the virtual mathematical keyboard as needed.
• Red Area: Open the drop-down menu to choose different mathematical or scientific functions.
  (You can also right-click in the text field to open this menu.)

• Click “Finish” to save your changes and return.
• Click “Cancel” to discard changes and exit.

How to implement advanced Math functions in Math Editor

nCr

1. Enter an uppercase C.
2. Use the virtual keyboard to add a superscript to the C.
3. Move the cursor to the end of the C.
4. Click the ↑ button at the bottom-left corner of the virtual keyboard — or press and hold the Shift key on your physical keyboard.
5. Use the virtual keyboard to add a subscript to the C.
6. You will now see the nCr format displayed in the Math Editor. Adjust the values of n and r as needed.

Calculus function

1. Right-click, or click on the triple-line icon to open the Tools menu.
2. Select Insert.
3. You’ll see options for Derivative, Integral, Sum, and Product functions.