Release of "Image Translation Feature (Beta)": Translate Banners and Graphs with Embedded Text in One Click | Dramatically Streamlining the Translation of Internal Documents Rich in Diagrams
Yaraku, Inc. has released a new "Image Translation Feature (Beta)" for its AI-powered translation tool, Yaraku Zen. This feature allows users to translate text within images, graphs, and diagrams in PowerPoint and PDF files while maintaining the original layout, significantly improving efficiency for document translation.
📋 Article Processing Timeline
- 📰 Published: March 30, 2026 at 23:32
- 🔍 Collected: March 30, 2026 at 22:56
Yaraku, Inc. (Headquarters: Shibuya-ku, Tokyo; Representative Director: Yu Sakanishi; hereinafter "Yaraku"), provider of the generative AI-powered translation support tool (CAT tool) "Yaraku Zen," implemented an update on March 28, 2026, to release the new "Image Translation Feature (Beta)." This allows for a complete workflow of layout preservation, text embedding, and post-editing within a single application.
By simply dragging and dropping PowerPoint and PDF files into Yaraku Zen, text within images and annotations in graphs are translated. The system generates translated documents while maintaining the layout as much as possible, and the text within the images can be post-edited directly within the Yaraku Zen interface.
This feature is expected to be utilized in the translation of various manuals, organizational charts, and presentation materials, realizing overwhelming efficiency in translation tasks for sectors such as manufacturing.
■ Feature Highlights: No Prompts Required. One-Click Conversion via Drag & Drop
The system automatically extracts and translates text within images, then automatically repositions and generates the translated text while preserving the original background and design.
There is no need for the time-consuming "prompt adjustments" or "fine-tuning of font sizes and layouts" required by conventional general-purpose generative AI. Simply drag and drop the file you wish to translate (pptx or pdf) onto the Yaraku Zen screen.
Furthermore, the generated text can be seamlessly post-edited within Yaraku Zen, allowing you to make manual adjustments, such as "make it a bit shorter."
[Image Translation Feature (Beta) Introduction Video]
■ Expected Use Cases and Suitable Industries/Sectors
This feature is ideal for the following industries and tasks that frequently use visual elements like diagrams and flowcharts and require multilingual deployment:
- Manufacturing (Automobiles, precision equipment, electronic devices, etc.)
Translation of product manuals, design blueprints, safety management manuals, and engineering project management documents.
- IT & Information Communications
Multilingualization of system configuration diagrams, network diagrams, organizational charts, etc.
- Medical & Life Sciences
Translation of business promotion materials and specialized flowcharts.
- Corporate Departments
Rapid sharing of internal regulations and presentation materials for overseas offices.
■ Background of Development: Strong Practical Needs from Global Companies
Until now, when translating text within images included in product manuals or overseas presentation materials, many companies faced the hassle of extracting text, processing it in translation tools, and then repositioning it using image editing software.
This feature was born from the requests of many companies expanding globally who wanted to "translate and edit images without breaking the layout," making it a feature directly linked to practical business needs.
■ 3-Step Technology Supporting "Image Translation"
This feature achieves complex processes with a single click through a proprietary processing system that links multiple advanced AI technologies.
1. Advanced Text Recognition and Extraction (OCR Processing)
Utilizing cloud-based image recognition AI, it precisely reads not only the text within the image but also the exact coordinates where the characters are placed, the layout (left-aligned, centered, etc.), and even the orientation of the text.
2. Translation and Alignment Processing
The extracted text is processed by the translation engine on Yaraku Zen. For sections where the structure must be maintained, such as list formats, a proprietary alignment function automatically analyzes and optimizes line breaks in the translated text to ensure the structure of the original document is not compromised.
3. Redrawing (Rendering) via Proprietary Algorithm
When returning the text to the image, a proprietary algorithm calculates the original text color, font weight, and background color from the image data. After erasing the original text by filling it with the extracted background color, the system considers the character width differences inherent in different languages (such as the difference in text volume between English and Japanese) and automatically adjusts the font size to ensure the translation fits naturally within the original frame.
■ Target Files and Usage Conditions
Target Files:
Images within PowerPoint (.pptx) and images within PDF files (when downloading as .pptx).
Target Plans: Company Plan and Enterprise Plan (*Separate free application is required for use).
■ Future Developments
While currently provided as a beta version, a full-scale release as a "Production Version (Public Version)" is planned within the year.
In future updates, we plan to support the translation of images embedded in Word and Excel files.
We will continue to contribute to service improvement and the support of multilingual communication.
By simply dragging and dropping PowerPoint and PDF files into Yaraku Zen, text within images and annotations in graphs are translated. The system generates translated documents while maintaining the layout as much as possible, and the text within the images can be post-edited directly within the Yaraku Zen interface.
This feature is expected to be utilized in the translation of various manuals, organizational charts, and presentation materials, realizing overwhelming efficiency in translation tasks for sectors such as manufacturing.
■ Feature Highlights: No Prompts Required. One-Click Conversion via Drag & Drop
The system automatically extracts and translates text within images, then automatically repositions and generates the translated text while preserving the original background and design.
There is no need for the time-consuming "prompt adjustments" or "fine-tuning of font sizes and layouts" required by conventional general-purpose generative AI. Simply drag and drop the file you wish to translate (pptx or pdf) onto the Yaraku Zen screen.
Furthermore, the generated text can be seamlessly post-edited within Yaraku Zen, allowing you to make manual adjustments, such as "make it a bit shorter."
[Image Translation Feature (Beta) Introduction Video]
■ Expected Use Cases and Suitable Industries/Sectors
This feature is ideal for the following industries and tasks that frequently use visual elements like diagrams and flowcharts and require multilingual deployment:
- Manufacturing (Automobiles, precision equipment, electronic devices, etc.)
Translation of product manuals, design blueprints, safety management manuals, and engineering project management documents.
- IT & Information Communications
Multilingualization of system configuration diagrams, network diagrams, organizational charts, etc.
- Medical & Life Sciences
Translation of business promotion materials and specialized flowcharts.
- Corporate Departments
Rapid sharing of internal regulations and presentation materials for overseas offices.
■ Background of Development: Strong Practical Needs from Global Companies
Until now, when translating text within images included in product manuals or overseas presentation materials, many companies faced the hassle of extracting text, processing it in translation tools, and then repositioning it using image editing software.
This feature was born from the requests of many companies expanding globally who wanted to "translate and edit images without breaking the layout," making it a feature directly linked to practical business needs.
■ 3-Step Technology Supporting "Image Translation"
This feature achieves complex processes with a single click through a proprietary processing system that links multiple advanced AI technologies.
1. Advanced Text Recognition and Extraction (OCR Processing)
Utilizing cloud-based image recognition AI, it precisely reads not only the text within the image but also the exact coordinates where the characters are placed, the layout (left-aligned, centered, etc.), and even the orientation of the text.
2. Translation and Alignment Processing
The extracted text is processed by the translation engine on Yaraku Zen. For sections where the structure must be maintained, such as list formats, a proprietary alignment function automatically analyzes and optimizes line breaks in the translated text to ensure the structure of the original document is not compromised.
3. Redrawing (Rendering) via Proprietary Algorithm
When returning the text to the image, a proprietary algorithm calculates the original text color, font weight, and background color from the image data. After erasing the original text by filling it with the extracted background color, the system considers the character width differences inherent in different languages (such as the difference in text volume between English and Japanese) and automatically adjusts the font size to ensure the translation fits naturally within the original frame.
■ Target Files and Usage Conditions
Target Files:
Images within PowerPoint (.pptx) and images within PDF files (when downloading as .pptx).
Target Plans: Company Plan and Enterprise Plan (*Separate free application is required for use).
■ Future Developments
While currently provided as a beta version, a full-scale release as a "Production Version (Public Version)" is planned within the year.
In future updates, we plan to support the translation of images embedded in Word and Excel files.
We will continue to contribute to service improvement and the support of multilingual communication.