I have recently observed a decline in the accuracy of the results. Much of the extracted content from images appears to be incorrect. For instance, I’ve noticed that whitespace is often missing, and numbers are occasionally misinterpreted as characters (e.g., "1" being mistaken for "l").
Has anyone else experienced similar issues? Are there any recommendations or methods to improve the reliability and accuracy of the extraction process? Perhaps is there any tricks? Should I maybe make the images larger? Crop images into several?
Kind regards,
I have recently observed a decline in the accuracy of the results. Much of the extracted content from images appears to be incorrect. For instance, I’ve noticed that whitespace is often missing, and numbers are occasionally misinterpreted as characters (e.g., "1" being mistaken for "l").
Has anyone else experienced similar issues? Are there any recommendations or methods to improve the reliability and accuracy of the extraction process? Perhaps is there any tricks? Should I maybe make the images larger? Crop images into several?
Kind regards,