Fingerprint door lock relies on the electronic components, particularly the fingerprint recognition device for its core security. The fingerprint recognition device determines the quality and safety of the product, based on factors such as recognition efficiency, accuracy, degree of difficulty in being hacked, and so on. Besides the quality hardware of the fingerprint recognition device, the algorithms used must also be of high standard to ensure maximum security.
In China, the algorithms used in fingerprint door knob can be broadly classified into three types:
The first type is image-based feature recognition algorithm. Its basic principle involves capturing images of the fingers on the sensors, and extracting the corresponding whole image. The identification occurs by comparing the match between the whole or partial images, and the algorithm can have self-learning function to address issues relating to small sensor area. This type of algorithm is derived from mobile phone fingerprint recognition algorithm and hence has low cost overall, making it suitable for small sensor applications.
The second type is feature point-based recognition algorithm. The principle used here involves extracting the unique features of the fingerprint image, after the fingerprint is pressed on the sensor and preprocessed. The algorithm involves comparison of the features on the image, hence it does not suffer from the issue of image variation, which can affect image-based recognition algorithms. However, they may have difficulty in recognition if the sensor area is too small or if the features are not clearly visible. As a result, this type of algorithm is not suitable for small sensors, like those used in mobile phones, and is relatively more expensive overall.
The third type is a hybrid algorithm, which combines both image-based and feature point-based recognition algorithms. This type of algorithm utilizes the advantages of both, to achieve better recognition ability. It has the potential to become the most popular algorithm in the future, as it offers good speed, efficiency, and cost-effectiveness for small sensor applications, as well as good security for larger sensor surfaces.
From the point of view of security, image-based algorithm is less secure, as images can be easily copied, and because the small sensors used in most devices often incorporate self-learning function, there is limited scope for detecting fraudulent activities using image comparison. Despite the lower cost of image-based algorithm, it is not used widely in fingerprint door locks, possibly due to security reason. Feature point-based algorithm, on the other hand, is very secure, as it uses unique features, and the sensor size is usually quite large, making it very difficult to tamper with. Hence it is the most commonly used algorithm in fingerprint door locks. The hybrid algorithm that combines both types of recognition algorithms offers good potential, in terms of balancing between cost-effectiveness and security.
Leave Your Message