Skip to main content

Image requirements

We recommend using the BlinkID Verify SDK, which will handle all the image requirements for you, and provide optimal images to the API with the smoothest UX.

However, if you're not currently using the Verify SDK, or you're just testing the API, you can use the following guidelines to ensure the best possible results for your images.

Image Formats​

​All endpoints accept JPEG and PNG images. The resolution of the image should be at least 1080p (1920×1080 px). Smaller resolutions are accepted, but won't provide optimal results. Larger resolutions are also accepted, but won't improve the scanning process significantly. We don't recommend going above 4k in any case, but even 4k is not necessary for most use cases.

We recommend using JPEGs, but make sure to use a high quality compression setting.

Positioning​

Ensure the document fills approximately 60% - 70% of the image and is the original, uncropped and unprocessed camera frame.

Image positioning

Unfortunately, BlinkID Verify can't detect features hidden from view. Ensure the document is fully visible.

Image occlusion

Framing and context​

Let the document image 'breathe' by leaving sufficient spacing on each side. This will help us better detect the edges of the document.

If you're using our liveness checks, make sure the captured images have enough physical real-world context, which is important for their accuracy.

Image framing

Lighting conditions​

Ensure good lighting conditions. Make sure the document is well-lit and avoid scanning against bright backgrounds.

Image lighting

Natural daylight produces best results. Avoid using your flashlight as it's likely to cause glare.

Image glare

Resolution​

Blurred images are difficult to read even for us humans. Our machine learning solution is fast and accurate, but still not superhuman.

Image blur

Capture both sides (if applicable)​

If you're capturing a multi-sided document (e.g. not a passport or similar), make sure to capture both sides. We're able to run a lot more checks, and provide a verification that's an order of magnitude more robust.

Conclusion​

It can be difficult to get users to meet these image requirements when performing self quality assurance. Repeated image failures cause immense frustration and quickly leads to abandonment. For these reasons, it's crucial to have an image acquisition process that gets it right on the first attempt. Try our client SDKs for an optimized user experience that maximizes first time success rates.

Last updated on