Biometric privacy protection based on combination of hiding and chaotic encryption

Abstract EN

With the expanded use of biometric systems, the safety of the biometric feature has become increasingly important.

When biometric images are transferred through unsafe channels or stored as raw data, they become at risk of theft, forgery and attack.

Data hiding is one of the main techniques of Privacy protection.

The goal of biometric data hiding is for adequate personal data is to be included in the cover of biometrics and to maintain recognition performance.

The paper idea introduces two levels of security based on hiding and encryption.

The eye image is segmented into two regions region of Interest (ROI) and Non-Region of Interest (NROI), the iris segmentation method depends on the Circular Hough Transform (CHT).

The privacy data is embedded with NROI and then reassemble the image with ROI (iris) to get the embedding image.

Then chaotic encryption is applied on the embedded image to get a high level of security.

The experimental results are tested using the CASIA1 data set.

The tests of hiding level are done using measurements such as PSNR and NC.

The results show that the suggested method gives a higher value of PSNR which means not destroy the cover image and the value of NC is (1) which means a perfect reconstruction of secret data.

The tests on encryption levels show good results using measurements such as histogram, correlation, and entropy.