Hi,
I need to extract the logo text of the source document and I'm currently using textDetection profile for extraction. Is there any option/setting to fine tune extraction of logo text?
Logo Text Detection
Was this article helpful?
0 out of 0 found this helpful
Comments
3 comments
Hi Vishnu,
Please post here your source document and performed settings to get more specific recommendations.
Hey helen,
I'm using the following settings,
language=English&profile=textExtraction&imageSource=auto&exportFormat=txt,xml&xml:writeFormatting=true
Thanks,
Vishnu
Hi Vishnu,
The settings are correct, but there are some issues with this logo which make text recognition difficult:
- The first word is typed by the decorative font and its letters are written together. For such cases, the word could be treated as a picture and recognition quality could be unsatisfying.
- Some words have the large gaps (almost the same width as the space character) between the letters, so they are recognized as several words.
- Some letters are extremely close to each other, so they could be recognized as a single character.
If it is possible, please consider using the other font for the logo. You can also find useful our Source Image Recommendations.
Please sign in to leave a comment.