Security certification is becoming for many applications, such as significant financial transactions. To date, PIN and password authentication is the most common method of access control eyes and face edges. Due to the finite length of the password, the security level is low and can be easily damaged. Adding a new dimension to the sensing mode driven state of art multimodal boundary face recognition system of the image-based solutions. It combines the active complex visual features extracted from the latest facial recognition model and uses a custom Recursive Convolution Neural Network (RCNN) issue facial authentications and selection capabilities to ensure the safety of face recognition. Echo function, because it is dependent on the geometry and material of the face, not disguised by the pictures and videos, such as multi-modal design is easy to image based face recognition system. Therefore, it does not require a special sensor to eliminate the extra cost solutions such as Face. The experiment, while it is possible to prevent the image / video spoofing, design has been shown to achieve a face recognition performance comparable to the face authentication based on state-of-the-art image. Proposed indicators, self-mixing classification rate, other single classification rate, and the equivalent rate in the classification algorithm is used to determine the optimal number of certified class action. Ready-made identified face eyes authentication app to analyze performance indicators, such as the use of false rejection users get the rate, false acceptance rate and the equivalent error rate. The results show that the algorithm to achieve a higher performance.