In what Way is the Minutiae Extraction Method an Efficient Fingerprinting Algorithm?

In what Way is the Minutiae Extraction Method an Efficient Fingerprinting Algorithm?

Review of the Literature

In the modern increasingly technological world, user-friendly and reliable person recognition, as well as verification is vital in many aspects of life. The behavioral and physical features, by way of biometrics, of an individual are critical elements to use in identifying and verifying people. As indicated by Ain, Shaukat, Nagra, and Raja (2018), fingerprinting has been applied as one of the most commonly used and popular biometric methods that have been used in a wide range of areas for the identification and verification of persons. A wide variety of research studies have been conducted on fingerprint classification issues, which has led to an increasing number of proposals for automatic fingerprint identification and verification being brought up. However, with the advancement in information technology, innovations are still coming up to provide more information on the method, which in turn makes this area of study to be active. While significant steps have been made in advancing automatic fingerprint identification approaches, a variety of design factors have limited the process of attaining the desired outcomes. The challenges in this view include the issues linked to the identification and verification of manual workers, unreliable minutia extraction algorithms, as well as the difficulty in the definition of fingerprint matches. Ain et al. (2018) also agree that some fingerprints are subject to high chances of degradation because of a variety of factors, including greasy, wet, wounded, scarred, or creased surfaces. Consequently, research and development in the area of approaches to enhance fingerprint quality and the ultimate recognition and verification performances are gaining more attention for researchers.

Evaluation of the Minutiae-Matching Algorithm

Currently, there are a wide variety of fingerprint matching algorithms that can be used, including minutiae matching, graph matching, transform feature matching, hybrid feature matching, as well as genetic algorithms (Chaudhari, Patnaik, & Patil, 2014). In real-world settings, for fingerprint identification and verification using the minute-matching algorithm, it is not always possible to achieve idle-quality images of fingerprints. It is also challenging to give homogeneous datasets. These are some of the main challenges that are associated with today’s fingerprint identification and verification methods.

                Ain et al. (2018) provide that the efficiency of the minutiae-based extraction technique can be improved in the future through the enhancement of features that would enhance the fingerprints’ verification capability. The minutiae-based algorithm has two main issues, including similarity and correspondence computation. With the similarity-based computation, the method extracts a 17-D vector from the matching fingerprint (Ain et al., 2018). Subsequently, a vector classifier is used to convert the outcome into a feature vector. Furthermore, for the correspondence-based computation, two descriptors are assigned to all the minutia points. Consequently, an alignment-based matching algorithm is applied to attain some level of correspondence of the obtained minutiae.

From another point of view, as proposed by Seng, Zhang, Fang, Zhang, and Chen (2018), the minute extraction method is less efficient as a fingerprinting algorithm. It is faced with issues such as difficulty in aligning fingerprints, image deformation, as well as low image resolution. Seng et al. (2018) provide that this approach remains inadequate, especially in terms of the achievement of the high-quality resolution and fingerprint extraction that will lead to the convenience and ability of the method to achieve effective fingerprint extraction, identification, and verification.

Wang, Gavrilova, Luo, and Rokne (2006) propose that the minutiae-based matching system can effectively achieve a degree of a binary decision of the matched and unmatched fingerprints. The study further provided the method has been the best fingerprint extraction and matching method, mainly because it has the highest matching, identification, and verification capability. On the other hand, Wang et al. (2006) propose that there are a variety of challenges for the minutiae-based approach that limits its efficiency, meaning that there is still room for improvement. Some of the challenges include that any missing and false minutiae should be considered. The system should ensure that the matching algorithm should have the ability to accommodate points in one image that do not have matching ones in other image sets. Wang et al. (2006) also argued that the minutiae-based method is considered by many sectors to be computationally expensive, which leads to additional costs being incurred for its use. It is also faced with the difficulty of high chances of nonlinear deformation of images. If such distortions are not corrected, then the extraction cannot achieve a perfect alignment.

Improvement of the Efficiency

                 Ain et al. (2018) propose a more efficient algorithm using the minutiae extraction approach while proposing that the improved system will have an improvement of 80% relative to the existing methods. The method consists of two main steps, including feature extraction and matching. In the first step, the fingerprint images are taken from the user and used as the input to a pre-processing technique (Ain et al., 2018). In this phase, image segmentation and enhancement are conducted with the image being binarized and morphological operations being conducted to extract the final minutiae. Furthermore, segmentation ensures that the image is partitioned into multiple segments that can be evaluated more. Consequently, the features are extracted and passed through a post-processing technique that involves the removal of false minutiae (Bansal, Sehgal, & Bedi, 2011). Ultimately, the remaining features are aligned and matched. Ain et al. (2018) argue that this method will ensure that the minutiae can be evaluated at the smallest possible level, which helps in matching fingerprints more effectively.

Concerning the findings by Ain et al. (2018), Chaudhari et al. (2014) make the finding that the minutiae extraction method requires to be improved to achieve efficiency. The study proposes a system that applies such stages as the preprocessing, minutiae extraction, and minutiae matching. It is important to segment the process into the normalization phase that helps in the standardization of the intensity values in the image through the identification of any unwanted features. The segmentation process involves the separation of the image’s foreground and background regions. All these processes are related to those proposed by Ain et al. (2018), and they ensure that the most delicate details of the fingerprint are captured in the process. Chaudhari et al. (2014) make the finding that the performance of the minutiae fingerprint extraction system is dependent on the fingerprint image’s quality. The Rutovitz Crossing Number (CN) is recommended to ensure that the minutiae extraction algorithm is able to detect all the features while differentiating between the true and false ones.

In light of the inefficiencies associated with the minutiae fingerprint extraction method, Seng et al. (2018) recommend a multi-view image recognition algorithm that will make more efficient. The principle involves a system where the finger is viewed through three cameras that have been set at different angles, which will ensure that the images are captured in three different views. The proposal is closely related to the findings of other studies, including Ain et al. (2018), because it recommends the use of different phases, including the preprocessing, feature extraction, matching, and ultimately the fusion of the images at different views. As per the findings by Seng (2018), the multi-view fingerprint extraction approach will help in the prevention of deformation of the fingerprint images. It also ensures that multiple images at different levels are acquired. These are beneficial effects in fingerprint recognition because they will achieve the legitimacy chances of a single image match. With the ability to accomplish the comparison between legitimate and illegitimate matching scores, there are high chances of meeting better image recognition outcomes. Furthermore, Seng (2018) proposes a system to recognize the matching of fingerprint direction, which makes it possible for the method to acquire information about different aspects of the fingerprints, including the field and texture of the neighborhood of the minutiae, especially in instances where there are few minutiae in the image. Such will help in solving a critical issue of the quality of the image, especially in the cases where the fingerprints being extracted are of a person who does manual work or have defective fingerprints (Bansal, Sehgal, & Bedi, 2011). The improvement will also help in the distinction and recognition of effective minutiae. Supporting comparison among algorithms ensures that the system plays a crucial role in ensuring that fingerprints are distinguished, which in turn helps in more effective fingerprint image matching and recognition.

Also, Wang et al. (2006) propose a topology-matching algorithm that can help in improving the minutiae-based extraction method. The strategy is based on three main ideas, including the use of Delaunay Triangle edges to achieve the matching index instead of using the whole minutiae triangles. The main objective of fingerprint verification is to determine if two images are from a similar finger or not. To effectively do this, the input image should be aligned with the template one that is represented by the patterns of the minutiae. A Delaunay Triangle edges are applied as the comparison index, which helps in attaining the transformation. Additionally, the method applies a different deformation model that helps in solving any elastic finger deformations that, in most instances, affect the quality of the process of verifying fingerprints. Valid fingerprint identification and verification method always solve the deformations. The framework proposed by Wang et al. (2006) helps in the quantification and modeling of the distortions in the fingerprints. Consequently, it uses the Radial Basis Functions (RBF) that provides a practical way to the challenge of modeling the deformations. The use of RBF has been examined in other sectors, especially in medical image matching and morphing. However, Wang et al. (2006) argue that it has not been tested with fingerprint verification. The third idea of this method is to achieve a higher quality of matching performance. Consequently, a maximum bipartite matching system is applied for the achievement of more accuracy.

Using the topology-matching algorithm tested under Wang et al. (2006) helps in the performance of fingerprint identification and verification with speed as compared to other matching techniques. The use of the triangle as the basis for comparison helps in easier and less costly matching processes. Consequently, it helps in enhancing the accuracy of the system. Furthermore, the use of RBF in modeling the deformations of the fingerprints and the improvement of the matching routine leads to more precision in matching and verification, as confirmed by Bansal, Sehgal, and Bedi (2011). The triangle also helps in ensuring that there are no challenges in matching fingerprint samples from the same person because the method can solve any non-linear distortions.

Future of the Study

Chaudhari et al. (2014) propose that future research should work towards improving the thinning process to prevent the breaking of ridges. Such an effect in the minutiae extraction algorithm might lead to a higher number of false minutiae, which will enhance accuracy. In addition, Seng (2018) proposes that further development in information technology might lead to more multi-faceted fingerprint extraction methods that will help in creating a better platform for differentiating between legitimate and illegitimate fingerprints. Developing more innovations to support the minutiae extraction method will help in the improvement of the quality, as well as the chances of matching the images. The application of the proposed improvements will help in ensuring that the minutiae-based extraction method achieves the identification and verification with more accuracy.

References

Ain, N., Shaukat, F., Nagra, A., & Raja, G. (2018). An Efficient Algorithm for Fingerprint Recognition using Minutiae Extraction. Pakistan Journal of Science, 70(2).

Bansal, R., Sehgal, P., & Bedi, P. (2011). Minutiae Extraction from Fingerprint Images – A Review. International Journal of Computer Science Issues, 8(5), 1694-0814.

Chaudhari, A., Patnaik, G., & Patil, S. (2014). Implementation of Minutiae Based Fingerprint Identification System Using Crossing Number Concept. Informatica Economică, 18(1).

Seng, D., Zhang, H., Fang, X., Zhang, X., & Chen, J. (2018). An improved fingerprint image matching and multi-view fingerprint recognition algorithm. Traitement du Signal, 35(3).

Wang, C., Gavrilova, M., Luo, Y., & Rokne, J. (2006). An efficient algorithm for fingerprint matching. Conference Paper.