Image Text Remover agent

  • Updated

Image Text Remover is an Optimizely Opal agent that removes overlaid text from an uploaded image and fills the cleared area with the surrounding background. 

  • Challenge – To remove captions, labels, or burnt-in subtitles from images, teams must manually edit them in dedicated design tools.
  • Agent outcome – The agent removes targeted text overlays from the image and paints each cleared area with the natural background.
  • Value – Marketing teams repurpose images without opening external editing software.

Required Optimizely products

None.

Install agent

Opal Administrators, Agent Builders, and Opal Users with the Add, edit, and install specialized agents  attribute for a custom role can add agents to their organization's Optimizely Opal instance. See Add users and set permissions.

Install the agent from the Opal Agent Directory.

  1. Go to Agents > Agent Directory.
  2. Select Image Text Remover.
  3. Click Install Agent to add it to your Opal instance.

    Screenshot of the Agent Directory with the Image Text Remover agent selected and Install agent highlighted.

Use the agent

In Opal Chat, enter @image_text_remover and provide the following information:

  • Image – The image to edit.
  • (Optional) Additional Instructions – Specific guidance on which text to remove. For example, remove only the bottom caption or there are two text blocks, remove both.

The agent removes the overlaid text, fills the cleared area with the surrounding background, and displays the edited image for review.

To use the Image Text Remover agent in a workflow agent, drag and drop it in your workflow. See Create a workflow agent.

Details

The Image Text Remover agent uses the following default settings. After you install the agent (see Install agent), customize these settings for your organization's needs. See Manage agents.

Input variables

The Image Text Remover agent uses the following inputs:

  • Image
  • (Optional) Additional Instructions

Tools

The Image Text Remover agent uses the following tools:

  • analyze_image_content
  • generate_or_edit_image
  • present_images

Additional details

The Image Text Remover agent uses the following default configuration:

  • Inference level – Balanced. The agent provides reliable answers with additional reasoning.
  • Files – None.
  • Output – Text.