Our Technology

We deploy a homomorphic encryption technique that allows us to process your information without deciphering it and the result remains usable. With the 3D image of your fingers, our sensors and software build a scan of your finger vein that will allow you to identify yourself in the following. 

The novelty of the innovation lies in the application of 3D technology to finger vein recognition in order to solve current biometric security issues in the healthcare sector while ensuring user privacy and providing access control as a key market application.

3D Finger Vein ID consists in a reliable biometric authentication system based on an innovative 3D reconstruction and matching method. The 3D Finger Vein ID technology comprises four main innovations:

The sensor captures the vascular image of veins, located about 3mm below the surface of the skin, using near infrared (NIR, a wavelength between 700 and 1000 nm) illumination, as haemoglobin absorbs the infrared light more effectively.

During the capture process, the NIR light is blocked by the haemoglobin. In the resulting image, the blood vessels appear as a dark pattern due to the high density of haemoglobin while the rest of the structure remains white, making possible to extract the vein pattern.

3D Finger Vein modelling and reconstruction protocol volumetric images of 3D structures are created computationally based on the images acquired by the 3D sensor. The 3D voxel data is then suitable for matching, segmentation, and recognition processes.

3D Finger Vein Biometric Recognition Protocol: It compares and matches the extracted 3D Finger Vein patterns against a pre-registered pattern database to recognize the individual. Generally, it involves three steps: normalization, feature extraction, and classification.

Once the ROI is extracted, images are enhanced by removing noise and/or lightening the vessels, a technique defined by the Contrast Limited Adaptive Histogram image that will be extracted, stored as a feature vector and processed for recognition. These features contain only relevant information such as orientation, length, and location. For instance, the vein pattern can be detected from the original image, reduced to 1 pixel-width (thinned) to form a skeleton and stored as a biometric feature. The vein recognition method will be the Minutiae Points Technique and for the skeleton extraction, we will use the multi-scale Gaussian Matched Filter.

A homomorphic encryption system: The system includes novel enrolment and authentication protocol based on threshold homomorphic encryption where the data is processed without decrypting it, so people's information remains confidential. This process eliminates the risk of data being stolen when a person's biometrics are being taken and verified. The communication algorithms ensure that if any data is stolen, it can be immediately tracked right back to the device from which the data was leaked.

For a more secure biometric identification including a very high protection of your private data, we elaborate a new technology named Global ID 3D Finger Vein.

Pr Lambert Sonna

CEO of Global ID
Lambert Sonna