Global Bilgi RPA (EN)

Menu

Cognitive

Cognitive

Aktiviteler

Color Detection

Analyzes the colors in an image and returns the most frequently used colors as a collection (list). Useful for extracting color statistics, filtering specific colors, or performing color-based analysis.

Usage Scenarios:

  • Detecting the dominant colors in an image

  • Analyzing brand colors in logos or product images

  • Performing color-matching or filtering

  • Checking whether certain colors exist in an image

  • Identifying visual quality or format inconsistencies

Parameters:

  • Image Path: Full path of the image file

  • Path Choose: Allows selecting the image file through UI

  • Top Count: Number of most frequent colors to return

  • Distinct: When true, returns only unique colors

  • Excluded Colors: Colors to exclude (RGB or Hex), separated by semicolons

Duckling Get

Identifies meaningful expressions such as dates, times, numbers, currency, units, quantities, or durations inside a text and converts them into structured data. Duckling is an open-source tool used for natural language processing (NLP).

Usage Scenarios:

  • Extracting monetary values, dates, or durations from free-text user input

  • Converting mixed or unstructured form text into meaningful parts

  • Using structured extraction in chatbot, form analysis, or document-processing workflows

  • Turning numeric or date-related expressions in natural language into structured values

Parameters:

  • Base Adress: Connection address of the Duckling server

  • Locale: Language and region format (e.g., en_GB, tr_TR)

  • Text: Raw text to be analyzed

Sample Output:
For the input “45$”, the following fields may be produced:

  • Body: “45$”

  • Value: “45”

  • Unit: “$”

  • Dimension: “amount-of-money”

  • Normalized Value: “45”

  • Normalized Unit: “USD”

  • Values: Collection of all detected values

Extract Images From PDF

Extracts embedded images from a PDF file and saves them to a specified target folder in the chosen format. The total number of extracted images is also returned as an output for use in workflow logic.

Usage Scenarios:

  • Extracting images from PDF reports or forms

  • Separating media from visually rich documents

  • Exporting signatures, graphics, or tables stored inside documents

  • Creating visual records for automated archiving or verification systems

Parameters:

  • Pdf Path: Full path of the PDF file (manual entry)

  • Pdf Path Choose: Selects the PDF file through UI

  • Image Path: Full path of the folder where extracted images will be saved

  • Image Path Choose: Selects the target folder through UI

  • Image Name: Base name for exported images (auto-numbered: image1.png, image2.png)

  • Pdf Password: Password for encrypted PDFs

  • Start Page: Starting page number

  • End Page: Ending page number

  • Save As Type: Output image format (bmp, tiff, jpeg, gif, png, vmf)

Fix Low Quality PDF

Improves the quality of a low-resolution or corrupted PDF file by generating a cleaner, fixed version. The processed PDF is saved to a specified path with a new file name.

Usage Scenarios:

  • Enhancing low-quality or blurry scanned PDFs

  • Preparing a damaged PDF for OCR or text extraction

  • Cleaning PDFs before data processing

  • Creating a readable version of auto-generated low-resolution PDFs

Parameters:

  • PDF Path: Full path of the input PDF

  • Fixed PDF Path: Full path where the improved PDF will be saved

  • PDF Path Choose: Select the input PDF through UI

  • Fixed PDF Path Choose: Select the output folder through UI

  • Fixed PDF File Name: Name of the improved PDF file

  • Delay Time: Wait time before execution (ms)

Image to Text

Extracts textual or numeric information from an image file using Optical Character Recognition (OCR) and outputs the result into a variable.

Usage Scenarios

  • Extracting text from invoices, receipts, or forms

  • Converting written content in images into digital data

  • Reading identifiers such as invoice numbers or national IDs

  • Capturing text from screenshots

  • Digitizing user inputs on image-based forms

Parameters

  • Path: Full file path of the image to be processed (e.g., “C:\New folder\sample.png”)

  • Path Choose: Selects the image file through the UI

Notes

  • The image must be clear and readable; blurry images reduce OCR accuracy

  • Font style, size, and contrast affect recognition quality

  • Large or high-resolution images may increase processing time

  • Only one of Path or Path Choose needs to be provided

Location From Image

Extracts GPS metadata from an image file and converts it into a readable address. The activity analyzes EXIF location data (latitude and longitude) and retrieves the corresponding physical address using a mapping service.

Usage Scenarios

  • Retrieve address information from photos taken on mobile devices

  • Verify fieldwork or inspection images with location data

  • Extract geographic details from report or social media images

  • Use location data in automated mapping workflows

Parameters

  • Path: Full file path of the image containing GPS data (e.g., “C:\Images\location.jpg”)

  • Path Choose: Select the image file manually via UI

  • Url: Direct URL of an online image

  • Api Key: Google Maps Geocoding API key

  • Delay Time: Wait time before execution (ms)

Notes

  • The image must contain GPS EXIF metadata

  • API key must be generated from Google Maps developer console

  • URL must be publicly accessible if supplied

  • If no GPS data exists, the result may return empty

  • Google API usage limits and quotas apply

Object Detection

Identifies objects within an image and returns structured information about each detected item. The output collection includes object name, confidence score, and bounding box coordinates (X, Y, Width, Height), enabling visual analysis and automated decision-making.

Usage Scenarios

  • Counting or validating objects in an image

  • Locating a specific object for interaction or automation

  • Verifying visual conditions (e.g., whether a coffee cup exists on a desk)

  • Extracting object-based insights for RPA workflows

Parameters

  • Path: Full file path of the image to be analyzed (e.g., “C:\resimler\gorsel.png”)

  • Path Choose: Interface option to manually select the image file

  • Delay Time: Waiting duration before execution (milliseconds)

Notes

  • Detection accuracy varies; check the returned confidence Score

  • Processing time increases with large or high-resolution images

  • Output is returned as a collection and can be looped, filtered, or mapped

  • Detected object names are returned in English (e.g., “laptop”, “book”, “chair”)

PDF To Excel

Converts a specified PDF file into an Excel workbook and saves the extracted table data in a structured spreadsheet format. Useful for transforming static PDFs into editable and analyzable Excel files.

Usage Scenarios

  • Converting invoice, receipt, or contract PDFs into Excel for reporting

  • Extracting tabular data from system-generated PDFs

  • Preparing structured datasets from scanned or downloaded documents

  • Automating financial or operational data transfer into Excel

Parameters

  • Pdf Path: Full file path of the PDF to be converted

  • Pdf Path Choose: Interface option to manually select the PDF file

  • Output Path: Folder path where the Excel file will be saved

  • Output Path Choose: Interface option to select output directory

  • Output File Name: Name of the generated Excel file

  • Pdf Password: Password for protected PDFs (optional)

  • Delay Time: Waiting duration before execution (milliseconds)

Notes

  • PDF should contain clearly structured tabular data for best results

  • Password-protected PDFs require the correct password to process

  • Output file may require formatting adjustments depending on layout

  • Large or image-based PDFs may increase processing time or require OCR

PDF To Text

This activity is used to read the text contained in a specified PDF file and assign it to a variable. It can operate on a page basis and extracts the content directly in text format.

Usage Scenarios:

  • Automatically reading data from PDF invoices, contracts, or forms
  • Extracting meaningful text from scanned PDFs
  • Pulling written text directly without requiring OCR
  • Processing or analyzing information from a specific page

Parameters:

  • Path: The full file path of the PDF document to be processed
  • Path Choose: Allows selecting the PDF file through the interface
  • Page: The page number from which the text will be extracted (e.g., “1”)
  • Language: Language selection
  • PDF Crop: Allows extracting text only from a specific region (coordinates) within the PDF
  • Delay Time: Waiting time before the operation starts (in milliseconds)

Points to Note:

  • The PDF file must contain text; OCR may be required for image-based content
  • If the page number is incorrect, the output may be empty
  • Content alignment may break in multi-column text documents
  • Processing time may increase for large files

Text Translate

This activity is used to translate a given text from a selected source language to a target language. It supports language selection, delay time, and output assignment.

Usage Scenarios

  • Translating form or email text into another language

  • Preparing multilingual reports or documents

  • Converting chatbot/user inputs into a target language

  • Translating OCR-extracted text into another language

Parameters

  • Text: The text to be translated. Example: “Ben İstanbul’da yaşıyorum.”

  • Source Language: The language of the input text. Example: “Turkish”

  • Target Language: The language to translate into. Example: “English”

  • Delay Time: Optional waiting time before execution (milliseconds).

Outputs

  • Translated Text: The translated text result.

Notes

  • The source and target languages must be selected correctly, otherwise translation may fail.

  • Longer texts may require a delay to avoid timeout issues.

  • Ensure UTF-8 compatibility when working with texts containing special characters.

Do you want to have Digital Employees?

Contact Us