Matching Error Rate Research Articles

Stereo Matching Method with Cost Volume Collaborative Filtering

Aiming at the problem of matching ambiguity and low disparity accuracy at the object boundary in stereo matching, a novel stereo matching algorithm with cost volume collaborative filtering is proposed. Firstly, for each pixel, two support windows are built, namely a local cross- support window as well as a global support window for the whole image. Secondly, a new adaptive weighted guide filter with a cross-support window as a kernel window is derived, and it is used to locally filter the cost volume. In addition, a minimum spanning tree is constructed in the whole image window, and then the minimum spanning tree filter is used to globally filter the cost volume. The collaborative filtering of cost volume is realized by fusing the filtering results of the local filter and global filter, so that each pixel can not only receive the support of the neighboring pixels in the local adaptive window, but can also receive the effective support of other pixels in the whole image, thus effectively eliminating the matching ambiguity in different texture regions while maintaining the disparity edges. The experimental results show that the average matching error rate of our method on the Middlebury stereo images is 3.17%. Compared with the other state-of-the-art methods, our method has higher robustness and matching accuracy, the generated disparity maps are smoother, and the disparity edges are better preserved.

Research on 3D virtual vision matching based on interactive color segmentation.

Given the prevalent issues surrounding accuracy and efficiency in contemporary stereo-matching algorithms, this research introduces an innovative image segmentation-based approach. The proposed methodology integrates residual and Swim Transformer modules into the established 3D Unet framework, yielding the Res-Swim-UNet image segmentation model. The algorithm estimates the disparateness of segmented outputs by employing regression techniques, culminating in a comprehensive disparity map. Experimental findings underscore the superiority of the proposed algorithm across all evaluated metrics. Specifically, the proposed network demonstrates marked improvements, with IoU and mPA enhancements of 2.9% and 162%, respectively. Notably, the average matching error rate of the algorithm registers at 2.02%, underscoring its efficacy in achieving precise stereoscopic matching. Moreover, the model's enhanced generalization capability and robustness underscore its potential for widespread applicability.

A secure fingerprint template generation mechanism using visual secret sharing with inverse halftoning

Fingerprints are the most popular and widely practiced biometric trait for human recognition and authentication. Due to the wide approval, reliable fingerprint template generation and secure saving of the generated templates are highly vital. Since fingers are permanently connected to the human body, loss of fingerprint data is irreversible. Cancelable fingerprint templates are used to overcome this problem. This paper introduces a novel cancelable fingerprint template generation mechanism using Visual Secret Sharing (VSS), data embedding, inverse halftoning, and super-resolution. During the fingerprint template generation, VSS shares with some hidden information are formulated as the secure cancelable template. Before authentication, the secret fingerprint image is reconstructed back from the VSS shares. The experimental results show that the proposed cancelable templates are simple, secure, and fulfill all the properties of the ideal cancelable templates, such as security, accuracy, non-invertibility, diversity, and revocability. The experimental analysis shows that the reconstructed fingerprint images are similar to the original fingerprints in terms of visual parameters and matching error rates.

Using the characteristic search algorithm in a library fingerprint identification system

As an important identification method, fingerprint recognition has a wide range of applications. To make the fingerprint recognition system of a library more efficient and secure, a recognition technology based on the characteristic search algorithm is proposed, and the performance of the algorithm is analysed. When a reasonable threshold is set, the matching error rate of the algorithm can be controlled at a lower level, and the algorithm can also ensure a higher fingerprint recognition accuracy and determine the overall accuracy. At the same time, three other identification algorithms of the same type are introduced: radio-frequency fingerprinting, convolutional neural network and local binary pattern. In a comparative experiment, it was found that the characteristic search algorithm model had the highest accuracy, with a value of 94.8%. When dealing with the same amount of fingerprint data, the recognition time of the algorithm model was the shortest. In addition, the area under the curve value corresponding to the receiver operating characteristic curve of the algorithm was the largest, and its value was 0.94. It is well known that the performance of the characteristic search algorithm is optimal and can effectively improve the operation efficiency of a library fingerprint identification system.

Multi-Layered Minutiae Extraction Based on Fusion-Attention for OCT Fingerprints

Minutiae extraction plays an important role in automated fingerprint identification systems. There are many minutiae extraction methods proposed for various fingerprints (rolled/slap/latent) in the past decades. However, for fingerprints by optical coherence tomography (OCT) imaging which have the essential property of multi-layered tissue structure, previous work may not be suitable. This paper is devoted to extract minutiae for OCT-based fingerprints based on the characteristics of OCT imaging and powerful deep neural networks. Specifically, we propose a multi-tissue-layered minutiae extraction network based on fusion-attention, involving a multi-tissue-layered feature extractor to fully explore the rich information in multiple subsurface fingerprints, and a one-stage detection framework to detect minutiae points. To demonstrate the effectiveness and robustness of the proposed method, a series of experiments have been carried out on the public OCT fingerprint benchmark dataset (https://github.com/FengLiu-0013/). Our method obtains the best results with an improvement of 8.3% in F1-score compared with state-of-the-art methods. The proposed method has also been evaluated for fingerprint identification using the extracted minutiae. Our method has achieved a matching error rate of 1.57%, with a relative reduction of 41% compared with state-of-the-art methods.

Real-time segmentation network for accurate weld detection in large weldments

Aiming at the defects of inaccurate weld extraction and high matching error rate in automatic welding system of large weldments currently. We propose a multi task detection model based on CNN architecture, which integrates the semantic segmentation technology required for weldment merging as well as the edge detection technology needed for weld matching. In particular, for the purpose of predicting smoother edges and welds, we carefully construct a new segment head, which adopts the sub-pixel convolution technology for up-sampling. Furthermore, a joint optimization loss function is explored to alleviate the imbalance of category distribution in large-scale weldment datasets. To verify the effectiveness of the model, abundant groups of data are collected for training and testing. The experimental results indicate that the proposed method has achieved the optimal trade-off between detection accuracy (83.35% mIoU, 95.15% F-score of welds and edges) as well as speed (74FPS) on a 2080Ti GPU compared with other state-of-the-arts, which greatly improves the robustness of the automatic welding system for large weldments.

Color Image Feature Matching Method Based on the Improved Firework Algorithm

An improved ORB feature point purification method has been proposed to address the problem of large feature point matching error and low image registration rate in the oriented FAST and rotated BRIEF (ORB) feature point color image matching algorithm. Pure virtual quaternions were used to represent the color image pixels in this method, and an improved FAST algorithm was used to detect the color feature points at first. The firework algorithm has been used to divide the detected feature points into key areas, auxiliary areas, and small influence areas, depending on the degree of attachment. The feature points of the key area and the subsidiary area are the required feature points. In order to further improve the efficiency of matching the color image feature points, the firework explosion radius formula and the explosion number formula in the firework algorithm have been improved, and an improved firework algorithm is proposed. This improved algorithm purifies the quaternion-represented color image feature points. For feature matching, the hamming distance was used. Experiments show that, when compared to the traditional ORB algorithm, the improved algorithm retains the key feature points of the color image with high accuracy, removes a large number of irrelevant feature points and noise points, and provides significantly higher accuracy and efficiency of color image matching.

Adaptive Superpixel-Based Disparity Estimation Algorithm Using Plane Information and Disparity Refining Mechanism in Stereo Matching

The motivation of this paper is to address the limitations of the conventional keypoint-based disparity estimation methods. Conventionally, disparity estimation is usually based on the local information of keypoints. However, keypoints may distribute sparsely in the smooth region, and keypoints with the same descriptors may appear in a symmetric pattern. Therefore, conventional keypoint-based disparity estimation methods may have limited performance in smooth and symmetric regions. The proposed algorithm is superpixel-based. Instead of performing keypoint matching, both keypoint and semiglobal information are applied to determine the disparity in the proposed algorithm. Since the local information of keypoints and the semi-global information of the superpixel are both applied, the accuracy of disparity estimation can be improved, especially for smooth and symmetric regions. Moreover, to address the non-uniform distribution problem of keypoints, a disparity refining mechanism based on the similarity and the distance of neighboring superpixels is applied to correct the disparity of the superpixel with no or few keypoints. The experiments show that the disparity map generated by the proposed algorithm has a lower matching error rate than that generated by other methods.

A Novel High-Resolution Fingerprint Representation Method

Current high-resolution fingerprints are almost entirely represented by sweat pores, resulting in information loss and weak robustness. This is caused by dividing fingerprint representation into pore detection and pore representation. This paper proposes a novel high-resolution fingerprint representation method for recognition based on only one fully convolutional network. By guiding the network to learn the most discriminative representation (i.e., sweat pores) and to reconstruct the original fingerprint image, pore maps and hidden features are provided simultaneously to represent the high-resolution fingerprint image for the subsequent direct matching process. The experimental results, which are evaluated on the public PolyU pore database, show that the best average <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$R_{T}$ </tex-math></inline-formula> , which is 94.88%, can be obtained using our proposed method. In addition, the recognition results show that our proposed method achieves matching equal error rates (EERs) of 5.55% and 1.27% on the PolyU DBI and DBII databases, respectively, which is better than other pore detection methods when using the same matching method. Compared with the latest DeepPoreID method which split fingerprint representation into pore extraction and pore representation, our approach improves the EER by up to 50.98% when evaluated on an in-house database, further demonstrating the outstanding generalization ability of the proposed method.

Unsupervised SAR and Optical Image Matching Using Siamese Domain Adaptation

Due to its highly complementary information about remote sensing, synthetic aperture radar (SAR) and optical imagery matching have drawn much attention in recent years. Compared with traditional methods, deep learning-based SAR-optical image matching models largely rely on supervision with ground truths, where the matching accuracy suffers because of unseen image domains. To mitigate loads in burdensome labeling tasks, transferring deep learning models trained with annotated source domains to nonannotated target domains has attracted great concern. Due to the domain gap, the difference between the source and target domains is likely to deteriorate the matching accuracy on target data if the training process is directly conducted without proper domain adaptation (DA). In this research, a Siamese DA (SDA) approach with a combined loss function is developed in the context of multimodality image matching. Then, a novel rotation/scale-invariant transformation module with regression modules is designed to extract rotation/scale-equivariant features. Finally, the causal inference-based self-learning method and the multiresolution histogram matching approach are employed to enhance the unsupervised matching performance. Experimental results on the RadarSat/Planet dataset and the Sentinel-1/2 dataset demonstrate that the developed model can achieve competitive matching performance with a low overlap ratio between domains and little data labeling. By alleviating the domain discrepancy, the developed model drastically reduces the average L2 score of the unsupervised matching from 9.576 to 0.658, while the less-than-one-pixel matching error rate is enhanced from 66.3% to 90.6%.

Template Extraction Algorithm using the Feature Analysis of Unique Patterns of Dog Nose Prints

This paper proposes a dog nose-print template-extraction algorithm using a feature analysis technique for unique nose-print patterns to recognize dogs. Dog recognition using nose prints begins with an accurate extraction of the unique shape, position, and direction feature information comprising a nose print. The proposed algorithm consisted of determining the object center of gravity, object size, and the distance information between the object center of gravity and inner approximation vertices for each unique nose-print pattern. Based on the features extracted from one nose-print shape, nose-print matching was performed by estimating the correlation between the position and direction of the neighboring nose-print patterns. Nose-print recognition test results using the template generated by the proposed method showed 100% matching rates when the registered and authenticated nose prints were identical. For different nose prints, the matching error rate was 4.59%, 1.14%, 0.12%, 0.02%, and 0% for a similarity of 10, 20, 30, 40, and 50% or more, respectively. The experimental results showed that the matching success rates of the same nose print were excellent. No false authentication occurred in the matching results for different nose prints when the similarity exceeded 50%.

Using citizen science in the photo-identification of adult individuals of an amphibian based on two facial skin features.

Among amphibians, adults have traditionally been identified in capture-mark-recapture studies using invasive marking techniques with associated ethical, cost and logistical considerations. However, species in this group may be strong candidates for photo-identification based on natural skin features that removes many of these concerns, with this technique opening up opportunities for citizen scientists to be involved in animal monitoring programs. We investigated the feasibility of using citizen science to distinguish between individuals of an Australian anuran (the sandpaper frog, Lechriodus fletcheri) based on a visual analysis of their natural skin features. We collected photographs of marked individuals in the field over three breeding seasons using a smartphone device. This photo-database was used to create an online survey to determine how easily members of the general public could photo-match individuals by a comparison of two facial skin features; black banding that runs horizontally above the tympanum and a background array of tubercles present in this region. Survey participants were provided with 30 closed, multiple choice questions in which they were asked to match separate images of a query frog from small image pools of potential candidate matches. Participants were consistently able to match individuals with a low matching error rate (mean ± SD of 26 ± 5) despite the relatively low quality of photographs taken from a smartphone device in the field, with most query frogs being matched by a majority of participants (mean ± SD of 86.02 ± 9.52%). These features were found to be unique and stable among adult males and females. Thus, photo-identification is likely to be a valid, non-invasive method for capture-mark-recapture for L. fletcheri, and likely many anurans that display similar facial skin features. This may become an important alternative to artificial marking techniques, with the challenges of manual photo-matching reduced by spreading workloads among members of the public that can be recruited online.

Low-error star pattern recognition based on point cloud matching algorithm*

Aiming at the problem that the matching rate of traditional star map recognition methods is not high enough, this paper proposes a star map recognition method based on the ICP point cloud matching algorithm based on feature point extraction. Feature points are selected through the change of the local normal vector of the point cloud, and then used Common (ICP) algorithm to realize more accurate registration of point clouds. In order to further reduce the error rate of star map matching, an improved ICP algorithm based on feature points is used to accurately register the point cloud. Finally, the matching conditions of several star maps are obtained through MATLAB tool simulation and the matching rate error rate is calculated. Common ICP algorithm The minimum matching error of is 14.224, and the minimum matching error of the improved ICP algorithm based on feature points is 14.0715. Obviously, the star map matching error rate of the ICP algorithm based on feature points is lower, which also verifies the superiority of the improved ICP algorithm.

4294 Patient Matching Errors and Associated Safety Events

OBJECTIVES/GOALS: Errors in patient matching could result in serious adverse safety events. Unlike publicized mix-ups by healthcare providers these errors are insidious and with increased data sharing, this is a growing concern in healthcare. The following project will examine patient matching errors and quantify their association with safety. METHODS/STUDY POPULATION: EHR systems perform matching out-of-the-box with unknown quality. Using matching processes outside the EMR, the rate at which matching errors are present was quantified and the erroneous records were flagged providing both comparative measures and data necessary to evaluate patient safety. To understand the relationship between matching and safety we will establish a percent of voluntarily reported safety events in our institution where a matching error existed during an encounter. Any safety events occurring for a flagged patient will be reviewed to determine if matching errors contributed to the safety problem. Not all safety events are reported so we will perform full chart review of a filtered list of medical records that have a higher likelihood of safety events. RESULTS/ANTICIPATED RESULTS: We were able to quantify matching errors, and the preliminary matching error rate is approximately 1%, representing over 700 patients. The work is in progress and we are beginning to determine the association between safety events and incorrect matching. Together these results will provide an incentive to identify errors, make corrections, and develop methods to achieve these objectives. The number of matching errors impacts patient care as well as business operations and is likely to have a negative financial impact on institutions with high error rates regardless of its relationship to safety. DISCUSSION/SIGNIFICANCE OF IMPACT: Patient matching is bundled with EHR software and institutions have little control over error rates, yet bear the liability for resulting clinical error. Institutions need to be able to identify undetected matching errors and any associated safety events and this project will provide that solution.

Improved Cost Computation and Adaptive Shape Guided Filter for Local Stereo Matching of Low Texture Stereo Images

Dense stereo matching has been widely used in photogrammetry and computer vision applications. Even though it has a long research history, dense stereo matching is still challenging for occluded, textureless and discontinuous regions. This paper proposed an efficient and effective matching cost measurement and an adaptive shape guided filter-based matching cost aggregation method to improve the stereo matching performance for large textureless regions. At first, an efficient matching cost function combining enhanced image gradient-based matching cost and improved census transform-based matching cost is introduced. This proposed matching cost function is robust against radiometric variations and textureless regions. Following this, an adaptive shape cross-based window is constructed for each pixel and a modified guided filter based on this adaptive shape window is implemented for cost aggregation. The final disparity map is obtained after disparity selection and multiple steps disparity refinement. Experiments were conducted on the Middlebury benchmark dataset to evaluate the effectiveness of the proposed cost measurement and cost aggregation strategy. The experimental results demonstrated that the average matching error rate on Middlebury standard image pairs is 9.40%. Compared with the traditional guided filter-based stereo matching method, the proposed method achieved a better matching result in textureless regions.

Map-matching algorithm based on the junction decision domain and the hidden Markov model.

Map-matching technology is a key and difficult technology in the development of vehicle navigation systems. Only by correctly identifying the road segment on which the vehicle is traveling can the navigation system make the right decision. At the same time, the complexity of the road network structure and a variety of error factors have introduced great challenges to map matching and have attracted the attention of many researchers as well. This paper analyzes various map-matching algorithms, determines that the key to the matching performance is the junction matching, performs an in-depth study on the junction-matching problem, and puts forward the junction decision domain model. The model mainly involves information regarding the width of the road segment, the angle between two road segments, the accuracy of GPS and the accuracy of the road network. In this paper, we use this model to improve the map-matching algorithm based on a hidden Markov model (HMM). The experimental results show that the improved matching algorithm can effectively reduce the error rate of junction matching and improve the matching performance of a navigation system.

Latent Fingerprint Identification Using Deep Learning Method

Digital fingerprint is one of the most consistent modalities in up to date biometrics and hence has been broadly studied and deploy in real applications. The accuracy of one Automatic Fingerprint Identification System (AFIS) largely depends on the quality of fingerprint samples, as it has an important impact on the degradation of the matching (comparison) error rates. This thesis generally focuses on the evaluation of biometric quality metrics and Fingerprint Quality Assessment (FQA), particularly in estimating the quality of gray-level latent fingerprint images or represented by minutiae set. By making a refined review of both biometric systems and relevant evaluation techniques, this contribute by the definition of a new evaluation or validation outline for estimating the performance of biometric quality metrics. It is defined to check the quality of latent fingerprint images by statistically measured parameters. In this work, an automatic Region-Of-Interest (ROI)-based latent fingerprint quality assessment technique is proposed by using deep learning. The first stage in our model uses deep learning, namely Region Convolutional Neural Network (R-CNN) to segment a latent fingerprint. In the second stage, feature vectors computed from the segmented latent fingerprint are used as input to a multi-class perceptron that predicts the value of the fingerprint. This proposed approach eliminates the need for manual ROI and feature markup by dormant examiners. Finally, experimental results on NIST SD27 show the effectiveness of our technique in latent fingerprint quality prediction

Stereo matching using census cost over cross window and segmentation-based disparity refinement

Stereo matching is a vital requirement for many applications, such as three-dimensional (3-D) reconstruction, robot navigation, object detection, and industrial measurement. To improve the practicability of stereo matching, a method using census cost over cross window and segmentation-based disparity refinement is proposed. First, a cross window is obtained using distance difference and intensity similarity in binocular images. Census cost over the cross window and color cost are combined as the matching cost, which is aggregated by the guided filter. Then, winner-takes-all strategy is used to calculate the initial disparities. Second, a graph-based segmentation method is combined with color and edge information to achieve moderate under-segmentation. The segmented regions are classified into reliable regions and unreliable regions by consistency checking. Finally, the two regions are optimized by plane fitting and propagation, respectively, to match the ambiguous pixels. The experimental results are on Middlebury Stereo Datasets, which show that the proposed method has good performance in occluded and discontinuous regions, and it obtains smoother disparity maps with a lower average matching error rate compared with other algorithms.

Research on anthropometry information obtaining system in garment MTM on the network mode

Purpose – The research is made in view of the anthropometry information obtaining problem in garment MTM on the network mode. The purpose of this paper is to obtain anthropometry information in a convenient and detailed way in garment MTM on the network mode. Design/methodology/approach – First of all, 24 main measurement sizes of 427 young females are collected to constitute the measurement database. The database is used as background data support of the system. The images are captured to simplify the way of inputting the anthropometry information to the system. Through the 2D feature sizes extracted from body image and the basic dimensions provided by customer input to the system, so that to gain the body sample which is closest to the customer body type through query matching in the database. The detailed anthropometry information of the closest sample is used to describe the customer. The human body measurement database and the technology of body image acquisition are used to extract the feature sizes to achieve obtaining the anthropometry information in a convenient and detailed way. Findings – Through query matching to the customer in a test, the body sample closest to the customer is gained, and the matching error rate is 0.0132. In the end, some customer samples are input to test the system, in order to verify the effectiveness of system functions. The matching error rates of five body types are gained all less than 0.006. The error is small, and the matching result is ideal. Research limitations/implications – The size of database established in the paper can be increased constantly in the future to obtain the more accurately matching result. Practical implications – The research of anthropometry information obtaining system in garment MTM on the network mode is the basis to achieve gaining the anthropometry information in a convenient and detailed way. Social implications – Applying the established system of human body measurement information acquisition in this paper, it can achieve to obtain the detailed measurement information of customer through a convenient way, combining the method of human body parameter model establishment in the existing research, it can achieve the complete network tailored mode with detailed measurement information acquisition and 3D virtual fitting functions. And it can provide the most convenient experience and the most ideal garment MTM effect to the customer. This mode can be forecast to be an ideal form of garment MTM on the network in the future. Originality/value – The anthropometry information obtaining system is the important part of garment MTM system on the network mode. It should be applied to the network mode and can obtain the detailed measurements for garment MTM. In this paper, the human body measurement database and the technology of body image acquisition are used in order to extract the feature size to obtain the anthropometry information in a convenient and detailed way.

Comparative Study of Multimodal Biometric Recognition by Fusion of Iris and Fingerprint

This research investigates the comparative performance from three different approaches for multimodal recognition of combined iris and fingerprints: classical sum rule, weighted sum rule, and fuzzy logic method. The scores from the different biometric traits of iris and fingerprint are fused at the matching score and the decision levels. The scores combination approach is used after normalization of both scores using the min-max rule. Our experimental results suggest that the fuzzy logic method for the matching scores combinations at the decision level is the best followed by the classical weighted sum rule and the classical sum rule in order. The performance evaluation of each method is reported in terms of matching time, error rates, and accuracy after doing exhaustive tests on the public CASIA-Iris databases V1 and V2 and the FVC 2004 fingerprint database. Experimental results prior to fusion and after fusion are presented followed by their comparison with related works in the current literature. The fusion by fuzzy logic decision mimics the human reasoning in a soft and simple way and gives enhanced results.