OCR Find Text

OCR Find Text

Type: ocr-find-text  ·  Package: OCR Activities v1.2.5  ·  Output: object

Searches for a text inside an image or screen region and returns its position (coordinates). Useful to locate an element by its label and then, for example, click near it — when there is no other way to find it.

How it works

Runs OCR on the source and searches for text word by word:

• Source: if you set path, it analyzes that image file and the region coordinates are ignored. Otherwise it captures the screen region x/y/width/height (all four are required in that case and need an interactive desktop).
• Word comparison: it returns the first recognized word that contains text, going through the lines top to bottom. Because it compares against single words, a text with spaces ("Amount due") is never found: search for the most distinctive word ("due"). "Total" does find the word Total:.
• Case: not case-sensitive by default (caseSensitive = false).
• Coordinates: in pixels, relative to the analyzed image. With a screen region, they are relative to the region’s corner: add the region’s x and y to get the screen position.
• This activity has no preprocessing.

Required parameters

Parameter	Type	Default	Description
text	string	—	Text to search for inside a word of the OCR result.

Optional parameters

Parameter	Type	Default	Description
path	string	—	Path to the image. If set, it is used instead of the screen region.
x	int	—	Left coordinate of the screen region (px). Required if there is no path.
y	int	—	Top coordinate of the screen region (px). Required if there is no path.
width	int	—	Width of the screen region (px). Required if there is no path; greater than zero.
height	int	—	Height of the screen region (px). Required if there is no path; greater than zero.
language	string	"eng"	Text language: Tesseract code (eng, spa, eng+spa) or regional tag (es-CO, en-US). See language.
caseSensitive	bool	false	Match upper and lower case exactly.

Output

Returns an object (dictionary):

Key	Type	Content
found	bool	true if the text was found.
x, y	number	Top-left corner of the word found (px).
width, height	number	Size of the word (px).
word	string	The full recognized word that contains text.
line	string	The full line where the word is.

If it is not found, it returns only { "found": false } (without the other keys); it is not an error.

Example

OCR Find Text   x = 200   y = 100   width = 800   height = 150   text = "Export"   → output: pos
If   condition = pos["found"]
 └─ then:
      Set Variable   name = centerX   value = 200 + num(pos["x"]) + num(pos["width"]) / 2
      Set Variable   name = centerY   value = 100 + num(pos["y"]) + num(pos["height"]) / 2
      Log   message = "Button on screen: " + centerX + ", " + centerY + " (line: " + pos["line"] + ")"

Check found first

Always read pos["found"] before using the coordinates: if the text does not appear, the other keys do not exist.

Common errors

Message	Cause and fix
'text' is required.	text is missing.
File not found.	The path does not exist on the robot’s machine.
'path' or screen coordinates 'x', 'y', 'width', 'height' are required.	There is no path and no x. Provide one of the two sources.
'y' is required for the screen region. (same for width, height)	x was given but another region coordinate is missing.
'width' and 'height' must be greater than zero.	The region size is 0 or negative.
Tesseract language '…' not found at '…'. Available: [eng, spa].	The language code is not installed.

Related activities

• OCR Recognize File — get all the text instead of a position.
• OCR Get Lines — all lines and words with their coordinates.