This person’s identification system is a state-of-the-art method that utilizes image processing algorithms to accurately identifies person’s true identity based on human retina’s unique vascular structure. This system has used digital retinal images known as fundus images for identification purpose. The proposed system has been used to be used as an efficient biometric system in high security areas in place of other biometric systems based on iris, finger print, face etc. as these can be forged easily.
This system is based on retinal images known as fundus images, acquired using fundus camera from different persons including both male and female. Relevant information has also been recorded for each person along with their unique labels. This recognition system is aimed to be deployed at a high-security area to allow the entrance of genuine employees and restrict the imposter ones. To achieve this, multiple samples of fundus images along with all related information is gathered from each employee (genuine) and recorded in the dataset. System is well trained on the given dataset and then tested for some new (unseen) samples to evaluate the reliability of the system.
System utilizes various image processing algorithms on the given dataset which include pro-processing, feature extraction and features matching based on features correlation metric. In order to check the accuracy, three publicly available retinal datasets i.e. DRIVE, STARE and VARIA are used. We have divided the results in two groups i.e. blood vessel segmentation and person recognition. In first group, we have tested the accuracy of proposed system for vascular segmentation as recognition depends on accurate extraction of blood vessels.
With this retinal recognition system, person identification can be done faster and more accurate, leading to better solutions to security issues. This system can be deployed in a real-world setting and used as a complementary tool to assist high-security areas in making better solutions. I would provide an accurate and reliable biometric system based on image processing and machine learning to my clients with experienced technical skills that can best serve my clients. If you are looking for an accurate and reliable image processing or ML-based image recognition system, look no further than me.