Latent fingerprint segmentation with adaptive total variation model

Latent fingerprint detection and segmentation play a critical role in image forensics for law enforcement. Being collected from crime scenes, a latent fingerprint is often mixed with other components such as structured noise or other fingerprints. Existing fingerprint recognition algorithms fail to work properly for latent fingerprint images, since they are mostly applicable under the assumption that the image is already properly segmented and there is no overlap between the target fingerprint and other components. In this work, we present a novel directional total variation (DTV) model to achieve effective latent fingerprint detection and segmentation. As compared with existing total variation models, the proposed DTV model differentiates itself by considering spatial-dependent texture orientations in the TV computation, which is particularly suitable for images with oriented textures. We demonstrate the superior performance of the proposed DTV technique using images from the NIST SD27 latent fingerprint database.

Latent fingerprints are captured from the fingerprint impressions left unintentionally at the surfaces of the crime scene. They are often used as an important evidence to identify criminals in law enforcement agencies. Different from the widely used plain and rolled fingerprints, the latent fingerprints are usually of poor quality consisting of complex background with a lot of nonfingerprint patterns and various noises. Latent fingerprint segmentation is an important image processing step to separate fingerprint foreground from background for more accurate and efficient feature extraction and matching. Traditional methods are usually based on the local features such as gray scale variance and gradients, which are sensitive to noise and cannot work well for latent images. This paper proposes a latent fingerprint segmentation method based on combination of ridge density and orientation consistency, which are global and local features of fingerprints, respectively. First, a texture image is obtained by decomposition of latent image with a total variation model. Second, we propose to detect the ridge segments from the texture image, and then compute the density of ridge segments and ridge orientation consistency to characterize the global and local fingerprint patterns. Finally, fingerprint segmentation is performed by combining the ridge density and orientation consistency for latent images. The proposed method has been evaluated on NIST SD27 latent fingerprint database. Experimental results and comparison demonstrate the promising performance of the proposed method.

Total variation has been used as a popular and effective image prior model in the regularization-based image processing fields. However, as the total variation model favors a piecewise constant solution, the processing result under high noise intensity in the flat regions of the image is often poor, and some “pseudo-edges” are produced. In this paper, we develop a regional spatially adaptive total variation (RSATV) model. Firstly, the spatial information is extracted based on each pixel, and then two filtering processes are respectively added to suppress the effect of “pseudo-edges”. After that, the spatial information weight is constructed and classified with kmeans clustering, and the regularization strength in each region is controlled by the clustering center value. The experimental results, on both simulated and real datasets, show that the proposed approach can effectively reduce the “pseudo-edges” of the total variation regularization in the flat regions, and maintain the partial smoothness of the highresolution image. More importantly, compared with the traditional pixel-based spatial information adaptive approach, the proposed region-based spatial information adaptive total variation model can better avoid the effect of noise on the spatial information extraction, and maintains robustness with changes in the noise intensity in the super-resolution process. Index Terms—Super-resolution, total variation, regional spatially adaptive, majorization-minimization

Latent fingerprints are the finger skin impressions which are left at the scene of a crime by accident. They are usually of poor quality with weak fingerprint ridge flows and various overlapping irrelevant patterns. It is still a challenging problem for automatic latent fingerprint processing and recognition. Latent fingerprint segmentation, which segments the fingerprint ridge area from complex backgrounds, is an important preprocessing step for latent fingerprint recognition. This paper proposes a latent fingerprint segmentation algorithm based on sparse representation. First, the total variation (TV) model is used to decompose a latent image into two components: texture and cartoon. The texture component, which contains the weak fingerprint ridge and valley structures, is used for further processing, while the cartoon component mainly consisting of the irrelevant information is discarded as noises. Then, we compute the sparse representation of the texture image against the dictionary constructed by a set of Gabor elementary functions. Since the sparse coefficients measure the weights of the basis atoms in fingerprint representation, an image quality measure is computed from the sparse coefficients, which evaluate how well the texture image can be sparsely reconstructed from the basis atoms. Finally, this image quality measure is used for fingerprint segmentation. We test the proposed method on the NIST SD27 latent fingerprint database. Experimental results and comparisons demonstrate the effectiveness of the proposed method.

Image enhancement plays an important role in biometric systems, this paper presented automatic latent fingerprint segmentation and matching. While considerable progress has made in both rolled and plain fingerprint image enhancement, latent fingerprint enhancement is a challenging problem due to the poor image quality of latent fingerprint with unclear ridge structures and various overlapping patterns, along with the presence of structured noise. Prior to latent fingerprint segmentation and feature extraction, latent fingerprint image enhancement is important to suppress various types of noise and to clarify the ridge structure. This paper reviews the current techniques used for latent fingerprint enhancement and presents a hybrid model which combines the edge directional total variation model (EDTV) and quality image enhancement with lost minutia reconstruction. The NIST SD27 database is used to test the performance of the proposed techniques with RMSE and PSNR. The proposed technique is effectively clarify input latent fingerprint images and eliminate noise in good, bad and ugly latent fingerprint images. A statistically significant difference, which focused on the mean lengths of PSNR and RMSE for different categories of latent fingerprint, images (good, bad and ugly). The proposed technique performs well for the good latent fingerprint images compare to bad and ugly images. Enhancement respectively presents RMSE averages of 0.018373, 0.022287, and 0.023199 for the good, bad and ugly image SD27 image set, as opposed to 82.99068, 81.39749, and 81.07826 for PSNR. The proposed enhancement technique improved the matching accuracy of latent fingerprint images by about 30%.

Latent fingerprints are the finger skin impressions left at the criminal scene unintentionally, which are important evidence for law enforcement agencies to identify criminals. Most of latent fingerprint images are of poor quality with unclear ridge structure and various non-fingerprint patterns. Segmentation is an important processing step to separate the fingerprint foreground from the background for more accurate and efficient feature extraction and identification. Traditional fingerprint segmentation methods are based on the information of gradients and local properties, which is sensitive to noise. This paper proposes a latent fingerprint segmentation algorithm based on linear density. First, a total variation (TV) image model is used to decompose a latent image into the cartoon and texture components. The texture component consisting of the latent fingerprint is used for further processing while the cartoon component is removed as noise. Second, we propose to detect a set of line segments from the texture image and compute the linear density map which can characterize the interleaved ridge and valley structure well. Finally, a segmentation mask is generated by thresholding the linear density map. The proposed method is tested on NIST SD27 latent fingerprint database. Experimental results and comparisons demonstrate the effectiveness of the proposed method.

The adaptive fidelity model and adaptive total variation(ATV) model are analyzed,and the strongpoint and disadvantage of the variational method models are compared according to experiments.The ATV model and texture preserving adaptive fidelity model are combined to deduce a gradient descent flow,and the result proved that it can remove noise effectively applying to remote sensing images,at the same time,the textures of the images are preserved.Finally,improved research tasks needed by remote sensing image noise removal based on partial differential equation are discussed.

Latent fingerprint matching assists for law enforcement agencies to identify criminals. Image enhancement plays an important role in automatic latent fingerprint segmentation and matching systems. Even-though sufficient progress done in both rolled and plain fingerprint images enhancement, latent fingerprint enhancement still a challenging problem and existing issue in the current research. This is due to the existence of poor quality images in latent fingerprint with unclear ridge structure and various overlapping patterns together with presence of structured noise. Prior to latent fingerprint segmentation and feature extraction, latent fingerprint image enhancement is necessary step to suppress different noises and improve the clarity of ridge structure. This paper reviews the current techniques used for the latent fingerprint enhancement. Thus, it presents hybrid model which is combination of edge directional total variation model (EDTV) and quality image enhancement with lost minutia reconstruction. NIST SD27 database has been used to test the proposed techniques and RMSE, PSNR to measure the performance. The result of the proposed technique shows enhancement of clarity of input latent fingerprint images and well de-noising of good, bad and ugly images of latent fingerprint. There is a statistically significant difference in the mean length of PSNR and RMSE for different categories of the latent fingerprint images (good, bad and ugly). It’s observed that the proposed technique performs well for the good latent fingerprint images compare to bad and ugly images. The result after enhancement present RMSE average 0.018373, 0.022287, and 0.023199 for the three different image categories available in SD27 data set good, bad and ugly images respectively while the PSNR average achieved 82.99068, 81.39749, and 81.07826 respectively. The proposed enhancement technique improved the matching accuracy of latent fingerprint images about 30{\%

Latent fingerprints have been used by law enforcement agencies to identify suspects for a century. However, because of poor image quality and complex background noise, latent fingerprints are routinely identified relying on features manually marked by human experts in practice. A large number of latent fingerprints can not be treated in time due to lacking well trained experts, highlighting the need for “lights out” (fully-automatic) systems. In this paper, we propose a systematic algorithm for latent fingerprint detection, segmentation, and orientation field estimation, without any manual markup. Multiple potential latent fingerprints are detected using a sequential pose estimation algorithm. Then, the full orientation field and confidence map of each detected fingerprint are estimated based on localized dictionaries lookup. Finally, the boundary of each latent fingerprint is delineated by analyzing its confidence map. Experiments on a multi-latent fingerprint database and the challenging NIST SD27 latent fingerprint database show the effectiveness of the proposed algorithm.

Latent fingerprint identification plays an important role for identifying and convicting criminals in law enforcement agencies. Latent fingerprint images are usually of poor quality with unclear ridge structure and various overlapping patterns. Although significant advances have been achieved on developing automated fingerprint identification system, it is still challenging to achieve reliable feature extraction and identification for latent fingerprints due to the poor image quality. Prior to feature extraction, fingerprint enhancement is necessary to suppress various noises, and improve the clarity of ridge structures in latent fingerprints. Motivated by the recent success of sparse representation in image denoising, this paper proposes a latent fingerprint enhancement algorithm by combining the total variation model and multiscale patch-based sparse representation. First, the total variation model is applied to decompose the latent fingerprint into cartoon and texture components. The cartoon component with most of the nonfingerprint patterns is removed as the structured noise, whereas the texture component consisting of the weak latent fingerprint is enhanced in the next stage. Second, we propose a multiscale patch-based sparse representation method for the enhancement of the texture component. Dictionaries are constructed with a set of Gabor elementary functions to capture the characteristics of fingerprint ridge structure, and multiscale patch-based sparse representation is iteratively applied to reconstruct high-quality fingerprint image. The proposed algorithm cannot only remove the overlapping structured noises, but also restore and enhance the corrupted ridge structures. In addition, we present an automatic method to segment the foreground of latent image with the sparse coefficients and orientation coherence. Experimental results and comparisons on NIST SD27 latent fingerprint database are presented to show the effectiveness of the proposed algorithm and its superiority over existing algorithms.

Latent fingerprints are one of the most important evidences used to identify criminals in the law enforcement and forensic agencies. Automated recognition of latent fingerprints is still challenging due to their poor image quality caused by unclear ridge structure and various overlapping patterns. Segmentation and enhancement are important to identify valid fingerprint regions, reduce the noise and improve the clarity of ridge structure for more accurate fingerprint recognition. In this paper, we propose a deep convolutional neural network architecture with the nested UNets for automatic segmentation and enhancement of latent fingerprints. First, to prepare training data, we synthetically generate the latent fingerprints and their segmentation and enhancement ground truth data for training. Then, a deep architecture of nested UNets is proposed to transform low-quality latent image into the segmentation mask and high-quality fingerprint through the pixels-to-pixels and end-to-end training. Finally, the test latent fingerprint is segmented and enhanced with the deep nested UNets to improve the image quality in one shot. The enhancement network is optimized by combining the local and global losses, which not only helps reconstruct the global structure, but also enhance the local ridge details of latent fingerprints. The proposed network can make use of multi-level feature maps in a pyramid way of nested UNets for segmentation and enhancement. Experimental results and comparison on NIST SD27 and IIITD-MOLF latent fingerprint databases demonstrate the promising performance of the proposed method.

Latent fingerprints are the impressions of ridges left on a crime scene due to unintentional touching of criminal’s fingers on different objects. These impressions of ridges have been used as vital evidences for identifying the criminals by law enforcement agencies. In this study, an attempt for automated segmentation of latent fingerprint has been made using K-means clustering. In proposed approach, Sobel filter and morphological operations have been used for background evaluation and image enhancement. Clustering is used to classify image data into k clusters to separate the foreground and background information. Mask has been generated on the basis of clustered data and is used to obtain the segmentation of latent fingerprints. Simulation results of the proposed algorithm show significant improvement in terms of missed detection rate (MDR) and false detection rate (FDR) using NIST SD-27 latent fingerprint database. Segmentation results reveal the MDR of 1.80, 4.75 and 7.80% and FDR of 17.85, 26.28 and 34.05% for good, bad and ugly quality of latent fingerprints, respectively. Moreover, visual segmentation reliability (VSR) is upto 90% for good quality images and varies in the range of 70–80% for bad quality latent fingerprints, whereas VSR for ugly fingerprints is in the range of 50–60%.

Latent fingerprint matching has played a critical role in identifying suspects and criminals. However, compared to rolled and plain fingerprint matching, latent identification accuracy is significantly lower due to complex background noise, poor ridge quality and overlapping structured noise in latent images. Accordingly, manual markup of various features (e.g., region of interest, singular points and minutiae) is typically necessary to extract reliable features from latents. To reduce this markup cost and to improve the consistency in feature markup, fully automatic and highly accurate ("lights-out" capability) latent matching algorithms are needed. In this paper, a dictionary-based approach is proposed for automatic latent segmentation and enhancement towards the goal of achieving "lights-out" latent identification systems. Given a latent fingerprint image, a total variation (TV) decomposition model with L1 fidelity regularization is used to remove piecewise-smooth background noise. The texture component image obtained from the decomposition of latent image is divided into overlapping patches. Ridge structure dictionary, which is learnt from a set of high quality ridge patches, is then used to restore ridge structure in these latent patches. The ridge quality of a patch, which is used for latent segmentation, is defined as the structural similarity between the patch and its reconstruction. Orientation and frequency fields, which are used for latent enhancement, are then extracted from the reconstructed patch. To balance robustness and accuracy, a coarse to fine strategy is proposed. Experimental results on two latent fingerprint databases (i.e., NIST SD27 and WVU DB) show that the proposed algorithm outperforms the state-of-the-art segmentation and enhancement algorithms and boosts the performance of a state-of-the-art commercial latent matcher.

A new image decomposition scheme, called the adaptive directional total variation (ADTV) model, is proposed to achieve effective segmentation and enhancement for latent fingerprint images in this work. The proposed model is inspired by the classical total variation models, but it differentiates itself by integrating two unique features of fingerprints; namely, scale and orientation. The proposed ADTV model decomposes a latent fingerprint image into two layers: cartoon and texture. The cartoon layer contains unwanted components (e.g., structured noise) while the texture layer mainly consists of the latent fingerprint. This cartoon-texture decomposition facilitates the process of segmentation, as the region of interest can be easily detected from the texture layer using traditional segmentation methods. The effectiveness of the proposed scheme is validated through experimental results on the entire NIST SD27 latent fingerprint database. The proposed scheme achieves accurate segmentation and enhancement results, leading to improved feature detection and latent matching performance.

We introduce a pre-enhancement algorithm to improve efficiency of the automatic fingerprint identification systems (AFIS) for latent fingerprint search. The proposed algorithm employs learning to construct a spectral dictionary from spectral responses of a Gabor filter bank in the frequency domain. Given an input latent fingerprint, the spectral dictionary yields a set of appropriate filters for each partitioning window of the entire latent fingerprint image. The proposed set of spectral filters helps improve and preserve highly-curved ridges in region around the singular point, while the other methods fail. The proposed method outperforms state-of-the-art algorithms in identification accuracy with the good and bad cases of the NIST SD27 latent fingerprint database.