🖥
🖥
🖥
🖥
AugeLab Studio Manual
English
Ask or search…
⌃K
Comment on page

Text Detection

Text Detection Node Documentation

The Text Detection node in AugeLab Studio is used to detect text regions in an input image. It utilizes the EAST (Efficient and Accurate Scene Text) text detection model to identify text regions and extract their bounding boxes.

Node Overview

The Text Detection node takes an input image and performs text detection on it. It identifies regions containing text and provides the coordinates of the bounding boxes around the detected text regions.

Node Interaction

  1. 1.
    Drag and drop the Text Detection node from the node library onto the canvas in AugeLab Studio.
  2. 2.
    Connect the input image to the node's input socket.
  3. 3.
    Adjust the confidence and NMS (Non-Maximum Suppression) thresholds in the node's properties panel.
  4. 4.
    Run the pipeline or execute the node to process the input image and detect text regions.
  5. 5.
    Retrieve the output image with text regions and the corresponding information from the node's output sockets.

Implementation Details

The Text Detection node uses the EAST (Efficient and Accurate Scene Text) text detection model to perform text detection on the input image. It follows these steps during its implementation:
  1. 1.
    Initialization:
    • The node loads the pre-trained EAST text detection model.
    • It sets the backend and target preferences for the model.
  2. 2.
    User Interface:
    • The node provides sliders to adjust the confidence and NMS thresholds.
    • The confidence threshold determines the minimum confidence score for considering a region as a text region.
    • The NMS threshold controls the overlap threshold for suppressing overlapping bounding boxes.
  3. 3.
    Text Detection:
    • The node receives an input image and applies the EAST text detection model to detect text regions.
    • It preprocesses the image and passes it through the model.
    • The model produces scores and geometry maps for each pixel in the image.
    • The node decodes the scores and geometry maps to extract text regions and their coordinates.
  4. 4.
    Output:
    • The node outputs the image with text regions highlighted using bounding boxes.
    • It also provides the reference point coordinates, reference rectangles, and the number of detected text regions as additional output information.

Usage

  1. 1.
    Drag and drop the Text Detection node from the node library onto the canvas in AugeLab Studio.
  2. 2.
    Connect the input image to the node's input socket.
  3. 3.
    Adjust the confidence and NMS thresholds in the node's properties panel.
  4. 4.
    Run the pipeline or execute the node to process the input image and detect text regions.
  5. 5.
    Retrieve the output image with text regions and the corresponding information from the node's output sockets.

Notes

  • The Text Detection node detects text regions in an input image using the EAST text detection model.
  • It provides sliders to adjust the confidence and NMS thresholds for controlling the detection results.
  • The node outputs the input image with text regions highlighted using bounding boxes.
  • It also provides the reference point coordinates, reference rectangles, and the number of detected text regions as additional output information.
  • The Text Detection node is useful for tasks such as optical character recognition (OCR), text extraction, and text analysis in images.
Function Blocks - AI Applications - Previous
Super Resolution Node
Last modified 4mo ago