Article
Computer Science, Information Systems
Diwakar Agarwal, Atul Bansal
Summary: Latent fingerprint identification, the most prevalent process used by the forensic community, is improved by proposing an algorithm that utilizes pores in addition to minutiae for matching. Experimental results demonstrate that the fusion of pores and minutiae significantly enhances the latent fingerprint recognition rate.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Chemistry, Multidisciplinary
Samy Bakheet, Shtwai Alsubai, Abdullah Alqahtani, Adel Binbusayyis
Summary: This paper presents an automated minutiae extraction and matching framework for accurately and quickly identifying and describing fingerprint minutiae. The proposed framework achieves comparable or superior performance to state-of-the-art methods in experimental evaluations, with an average equal error rate (EER) value of 2.01%.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Theory & Methods
Yulin Feng, Ajay Kumar
Summary: This paper proposes a new method for fast and accurate minutiae detection in fingerprint images. It uses a lightweight pixelwise local dilated neural network to extract local features and a patch-wise global neural network to recover the global features. The proposed method consolidates local and global fingerprint features, accurately localizes minutiae positions, and improves the minutiae detection accuracy and fingerprint matching accuracy. The method also has potential applications in other key points detection tasks.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2023)
Article
Environmental Sciences
Runlin Cai, Guanwei Shang
Summary: The proposed flexible 3-D Gabor features fusion (F3DGF) approach effectively utilizes both parts of Gabor features by introducing the phase-induced 3-D Gabor feature, leading to improved classification performance in hyperspectral image tasks.
JOURNAL OF APPLIED REMOTE SENSING
(2021)
Article
Computer Science, Information Systems
Abdul Wahab Muzaffar, Farhan Riaz, Tarik Abuain, Waleed Abdel Karim Abu-Ain, Farhan Hussain, Muhammad Umar Farooq, Muhammad Ajmal Azad
Summary: In this study, a novel rotation and scale invariant texture classification method based on Gabor filters is proposed. These filters are designed to capture the visual content of the images by their impulse responses, which are sensitive to rotation and scaling. The proposed method rearranges the filter responses and calculates patterns after binarizing the responses based on a specific threshold. The effectiveness of the proposed feature extraction method is demonstrated through experiments on famous texture datasets, and it is shown to be more robust to noise compared to other state-of-the-art methods considered in the study.
Article
Computer Science, Artificial Intelligence
Ting-Wei Shen, Mao-Hsiu Hsu, Chun-Hsu Shen, Wen-Fang Wu, Yu-Chiao Lu, Chia-Chun Chu
Summary: In this article, an algorithm based on the differential values of grayscale intensity for fingerprint orientation field (OF) estimation is introduced. The accuracy and reliability of the algorithm are examined by applying it to fingerprint images processed with Gaussian blurring and Gaussian white noise. Experimental results show that the proposed algorithm has higher OF estimation reliability than gradient-based and power spectral density (PSD)-based methods, especially in noisy fingerprint images, with improvements of 6.46% and 32.93% over the gradient-based and PSD-based methods, respectively.
PEERJ COMPUTER SCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Mohammad S. Khrisat, Rushdi S. Abu Zneit, Hatim Ghazi Zaini, Ziad A. Alqadi
Summary: This paper explores popular methods used to extract fingerprint features, testing and analyzing their efficiency, accuracy, flexibility, and sensitivity to image rotation. A detailed comparison analysis of MLBP, K_means, WPT, and Minutiae methods is conducted using multiple color images in various rotation modes to ensure the stability of image features.
TRAITEMENT DU SIGNAL
(2021)
Article
Engineering, Civil
Xinghao Ding, Fujin He, Zhirui Lin, Yu Wang, Huimin Guo, Yue Huang
Summary: Crowd counting is crucial in intelligent transportation systems, but challenges like occlusion, perspective distortion, and complex backgrounds make it difficult to achieve accuracy. This study introduces a novel CNN model and a new evaluation method for measuring density map accuracy, outperforming existing methods. Evaluation on cross-scene datasets shows promising performance of the proposed method.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Multidisciplinary Sciences
H. Mohamed Khan, P. Venkadesh
Summary: This study proposes a cross-sensor based authentication method using region-based minutiae count feature to mitigate the cross-sensor issue in fingerprint biometrics. By combining multiple fingerprint features and using the weighted sum rule, the reliability and robustness of fingerprint biometrics are improved. Experimental results show impressive performance with fingerprint accuracy reaching 97.87% and 98.50% for two datasets.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2023)
Article
Engineering, Biomedical
Rubab Mehboob, Hassan Dawood
Summary: This paper proposes an improved feature descriptor called Distinctively Encoded Histogram of Fingerprint Features (DEHFF) for live fingerprint detection. DEHFF incorporates ridge-valley contrast and phase of fingerprints, and uses Gabor filters to extract ridge contour information. By quantizing the extracted features into predefined intervals and integrating them, the average classification error rate is effectively reduced.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Article
Computer Science, Information Systems
Jitesh Pradhan, Arup Kumar Pal, Haider Banka
Summary: By decomposing images into object regions and non-object regions, and extracting corresponding salient and texture features, the accuracy and efficiency of image retrieval can be improved.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2022)
Article
Computer Science, Artificial Intelligence
Yingying Chen, Xiaowei Yang
Summary: In the field of support vector machines, online random feature map algorithms play a crucial role in large-scale nonlinear classification problems. However, the existing methods have shortcomings that can be overcome by the proposed random features based online adaptive kernel learning (RF-OAK) method, which shows superior performance through theoretical analysis and experiments.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Artificial Intelligence
Farhat Afza, Muhammad Attique Khan, Muhammad Sharif, Seifedine Kadry, Gunasekaran Manogaran, Tanzila Saba, Imran Ashraf, Robertas Damasevicius
Summary: This article proposes an action recognition technique based on features fusion and best feature selection, achieving high recognition rates on multiple famous datasets. The experimental results demonstrate the superior performance of the proposed scheme compared to listed methods.
IMAGE AND VISION COMPUTING
(2021)
Article
Chemistry, Analytical
Wen Liu, Changyan Qin, Zhongliang Deng, Haoyue Jiang
Summary: In this paper, a WiFi and visual fingerprint localization model based on low-rank fusion (LRF-WiVi) is proposed, which effectively utilizes the complementarity of heterogeneous signals and achieves superior positioning performance.
Article
Computer Science, Artificial Intelligence
Yicheng Li, Yingfeng Cai, Reza Malekian, Hai Wang, Miguel Angel Sotelo, Zhixiong Li
Summary: A new map creation method based on multi-sensor fusion technique was proposed for high-accuracy localization in semi-open scenarios for intelligent vehicles. The method employed road scenario fingerprint (RSF) to fuse visual features, 3D data, and trajectories, with experimental tests showing promising results for IV localization.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)