OCR Input Images - Best practices
Here you'll find guidelines on the recommended input image specifications for the OCR API. Following these guidelines will result in accurate OCR analysis and reduce processing time and bandwidth usage.
When using OCR analysis with Signzy's APIs, it is important to ensure that the minimum resolution of the input document image is 1024 x 768. OCR requires a higher resolution to detect characters accurately. Smaller resolutions may result in lower accuracy, while larger resolutions may increase processing time and bandwidth usage without providing significant benefits in accuracy.
When using OCR analysis with Signzy's APIs, it is important to note that the recommended image size should not exceed 20M pixels (length x width), and each image should be less than 2 MB. Using larger image sizes may increase processing time and bandwidth usage without providing significant benefits in accuracy.
It is important to keep in mind that this recommendation is based on the native resolution of the camera. Generating a high-resolution image from a low-resolution image does not produce the results needed for OCR due to artifacts generated by the up-sampling of the image.
The following guidelines are recommended be followed to ensure optimal image quality for OCR analysis:
Guideline | Description |
Image Composition | Use an image that is not obscured or tightly cropped. Cropped or partial images may lead to incomplete or incorrect OCR results, so ensure that the entire document is visible in the image. |
Color Images | It is recommended to use color images for OCR analysis. OCR algorithms perform better on color images than on black and white or grayscale images. If a color image is not available, make sure the image is of high contrast and has clear text. |
Lighting | Use images with flat lighting on the document, as opposed to varied lightings such as shadows and flashlights. Even lighting across the document helps OCR algorithms to recognize characters accurately. |
Image Quality | Use images that are bright and sharp. Avoid using images that may be blurry due to subject and camera motion as much as possible. Blurry images may cause OCR algorithms to misinterpret characters, resulting in incorrect OCR results. |
Blockage | Avoid items that block the document text, such as fingers, watermarks, stamps, or other objects. Such objects may interfere with OCR analysis, resulting in incorrect OCR results. |
Background | Avoid documents containing background texts like newspapers, keyboards, or any other items that may interfere with the OCR analysis process. |
Orientation | Make sure the orientation of the document is correct before scanning or uploading the image for OCR analysis. OCR algorithms work better on correctly oriented images, and it helps to avoid incorrect OCR results. |
Additional instructions for each of the APIs are provided in their respective documentation pages.
Following these guidelines will ensure that your input images meet the recommended specifications for OCR analysis, resulting in accurate and efficient processing. If you have any questions or require further assistance, please refer to the respective sections or contact our support team.