Sclera Segmentation Research Articles

SRU-Net: a novel spatiotemporal attention network for sclera segmentation and recognition

SRU-Net: a novel spatiotemporal attention network for sclera segmentation and recognition

Sclera Segmentation and Recognition for Spectacle Dataset

Sclera, a connective tissue enveloping the eye, emerges as a novel biometric recognition method for human identification. The composition of blood vessels in the sclera proves ideal for biometric use—visible with ease, stable over time, and unique to each individual. This paper proposes sclera segmentation and recognition techniques tailored for individuals wearing spectacles. The SSV (Spectacle Sclera Vision) Dataset was meticulously created to address challenges introduced by eyewear, including reflections, distortions, and illumination variations. The study explores the unique characteristics of the sclera region, presenting a comparative analysis of traditional and neural network-based segmentation and recognition methods on the SSV Dataset. Notably, Linear SVC outperforms CNN in recognition, and UNET demonstrates superior sclera segmentation compared to OTSU. The findings provide a foundation for potential advancements in developing robust multi-class classification models for sclera biometrics in real-world scenarios. Future work involves further analysis, scalability testing, and exploration of diverse applications in ocular health and security systems.

Sclera-TransFuse: Fusing Vision Transformer and CNN for Accurate Sclera Segmentation and Recognition

Sclera-TransFuse: Fusing Vision Transformer and CNN for Accurate Sclera Segmentation and Recognition

Crowdsourcing image segmentation for deep learning: integrated platform for citizen science, paid microtask, and gamification.

Segmentation is crucial in medical imaging. Deep learning based on convolutional neural networks showed promising results. However, the absence of large-scale datasets and a high degree of inter- and intra-observer variations pose a bottleneck. Crowdsourcing might be an alternative, as many non-experts provide references. We aim to compare different types of crowdsourcing for medical image segmentation. We develop a crowdsourcing platform that integrates citizen science (incentive: participating in the research), paid microtask (incentive: financial reward), and gamification (incentive: entertainment). For evaluation, we choose the use case of sclera segmentation in fundus images as a proof-of-concept and analyze the accuracy of crowdsourced masks and the generalization of learning models trained with crowdsourced masks. The developed platform is suited for the different types of crowdsourcing and offers an easy and intuitive way to implement crowdsourcing studies. Regarding the proof-of-concept study, citizen science, paid microtask, and gamification yield a median F-score of 82.2, 69.4, and 69.3 % compared to expert-labeled ground truth, respectively. Generating consensus masks improves the gamification masks (78.3 %). Despite the small training data (50 images), deep learning reaches median F-scores of 80.0, 73.5, and 76.5 % for citizen science, paid microtask, and gamification, respectively, indicating sufficient generalizability. As the platform has proven useful, we aim to make it available as open-source software for other researchers.

Ensemble convolutional neural networks and transformer-based segmentation methods for achieving accurate sclera segmentation in eye images

Ensemble convolutional neural networks and transformer-based segmentation methods for achieving accurate sclera segmentation in eye images

MetaScleraSeg: an effective meta-learning framework for generalized sclera segmentation

MetaScleraSeg: an effective meta-learning framework for generalized sclera segmentation

Exploring Bias in Sclera Segmentation Models: A Group Evaluation Approach

Bias and fairness of biometric algorithms have been key topics of research in recent years, mainly due to the societal, legal and ethical implications of potentially unfair decisions made by automated decision-making models. A considerable amount of work has been done on this topic across different biometric modalities, aiming at better understanding the main sources of algorithmic bias or devising mitigation measures. In this work, we contribute to these efforts and present the first study investigating bias and fairness of sclera segmentation models. Although sclera segmentation techniques represent a key component of sclera-based biometric systems with a considerable impact on the overall recognition performance, the presence of different types of biases in sclera segmentation methods is still underexplored. To address this limitation, we describe the results of a group evaluation effort (involving seven research groups), organized to explore the performance of recent sclera segmentation models within a common experimental framework and study performance differences (and bias), originating from various demographic as well as environmental factors. Using five diverse datasets, we analyze seven independently developed sclera segmentation models in different experimental configurations. The results of our experiments suggest that there are significant differences in the overall segmentation performance across the seven models and that among the considered factors, ethnicity appears to be the biggest cause of bias. Additionally, we observe that training with representative and balanced data does not necessarily lead to less biased results. Finally, we find that in general there appears to be a negative correlation between the amount of bias observed (due to eye color, ethnicity and acquisition device) and the overall segmentation performance, suggesting that advances in the field of semantic segmentation may also help with mitigating bias.

Anaemia detection based on sclera and blood vessel colour estimation

Anaemia is a global public health problem with major consequences for human health. Noninvasive methods must be investigated to determine a sick person's anaemic status or conduct screening campaigns, especially in resource-constrained areas of the earth. This study aims to prove that the colour of the sclera and scleral blood vessels extracted from digital images of the eye can be used to check the anaemic status of a person. To date, we have not found in the literature other studies that have attempted this promising approach. We propose a novel pipeline for anaemia estimation consisting of three main contributions: a sclera segmentation algorithm applied to near-taken digital photos of the eye, a vessel extraction algorithm, and a classifier to predict the anaemic status of a person vs normal controls. This study was based on the public dataset Eyes-defy-anaemia, which contains 218 eye pictures taken with a special device that removes any ambient light influence. Very interesting results have been achieved for the sclera segmentation task with good precision (88.53), recall (82.53) and F1 (84.10). The colour features and haemoglobin value appear to be well related allowing us to obtain an F2 score in the anaemia detection task of 86.4% using colour features from the whole sclera and of 83.8% using only vessels’ colour features.

Sclera biometrics in restricted and unrestricted environment with cross dataset evaluation

Research on ocular biometric systems in unrestricted environment has grown steadily over the last decade. Though sclera recognition is widely investigated as an alternative or supplementary ocular biometric modality, most of the work has been done in restricted environment. So there are several unanswered questions like, how a model trained in a restricted environment performs in an unrestricted scenario, or whether the performance improves if it is trained in a combined environment. In this work, we aim to study sclera biometrics in multiple environments and answer these questions. In the process, we have worked on all components in the sclera biometric pipeline, viz., sclera segmentation, vessel extraction, gaze detection and sclera recognition. We have trained and tested the deep models proposed for the tasks with high-quality datasets prepared in restricted environment as well as mobile datasets prepared in unrestricted environment. We have also trained the models with images from two or more datasets. This helps us to ascertain how the images captured in diverse environments can be used collectively for the training purpose. We have used two high-quality datasets SBVPI, MASD and two mobile datasets MSD, MASDUM for our work. MASDUM is the proposed mobile dataset prepared in unrestricted environment. It is designed specifically with research of sclera biometrics in mind. It contains a total of 1055 eye images along with 1055 sclera-segmented and 68 sclera-vessel-mark-up images. Since there is a severe lack of vessel-annotated datasets in the field, MASDUM dataset hopefully caters the need of researchers. State-of-the-art results are achieved for sclera segmentation on the datasets SBVPI, MASD, MSD and MASDUM with F1-scores of 96.3, 97.3, 87.9 and 91.2 respectively. State-of-the-art results are also achieved for sclera recognition with respective EERs of 0.132, 0.079, 0.086 and 0.080. We also present in-depth analysis on the novel idea of Dual-output variant of recognition model DeepR. This variant directly produces the matching score between two vessel-extracted binary images and it shows better results by largely reducing FAR.

Automated Mathematical Algorithm for Quantitative Measurement of Strabismus Based on Photographs of Nine Cardinal Gaze Positions.

This study presents an automated algorithm that measures ocular deviation quantitatively using photographs of the nine cardinal points of gaze by means of deep learning (DL) and image processing techniques. Photographs were collected from patients with strabismus. The images were used as inputs for the DL segmentation models that segmented the sclerae and limbi. Subsequently, the images were registered for the mathematical algorithm. Two-dimensional sclera and limbus were modeled, and the corneal light reflex points of the primary gaze images were determined. Limbus recognition was performed to measure the pixel-wise distance between the corneal reflex point and limbus center. The segmentation models exhibited high performance, with 96.88% dice similarity coefficient (DSC) for the sclera segmentation and 95.71% DSC for the limbus segmentation. The mathematical algorithm was tested on two cranial nerve palsy patients to evaluate its ability to measure and compare ocular deviation in different directions. These results were consistent with the symptoms of such disorders. This algorithm successfully measured the distance of ocular deviation in patients with strabismus. With complementation in the dimension calculations, we expect that this algorithm can be used further in clinical settings to diagnose and measure strabismus at a low cost.

An efficient deep sclera recognition framework with novel sclera segmentation, vessel extraction and gaze detection

Sclera recognition is a promising ocular biometric modality because of contact-less, gaze-independent image acquisition in visible light. Moreover, it is unaffected even if the subjects are wearing contact lenses in eyes. However, it is a difficult task because several steps are required, each of which must be performed accurately and efficiently. In this work, sclera recognition is performed in the following steps, namely, segmentation of sclera region, extraction of sclera vasculature pattern, detection of gaze direction and finally comparison of two vasculature patterns for matching and recognition. The proposed segmentation model DSeg is based on well-known deep learning model UNet and reduces model complexity by creating a Knowledge Base of sclera and non-sclera colors. DSeg is a lightweight and environment-friendly model, which outperforms UNet in terms of speed, efficiency and accuracy. Two rule-based unsupervised vessel extraction methods require prior sclera segmentation and exhibit competing recognition performance to a supervised deep model for vessel extraction, which does not require prior sclera segmentation. A novel deep recognition model is proposed which compares two vessel structures taking into account their affine-transformation, and produces a single Boolean output to decide whether the structures match or not. The model does not require post logic in the matching process. The model is further improved to detect errors in prediction. We achieve best recognition rates with low false-acceptance-rates for two sets of training and validation, using the publicly available dataset SBVPI and the best achieved AUC score is 0.98.

Automatic ocular version evaluation in images using random forest

The version exam is a test performed by specialists in ophthalmology to detect restrictions, paralysis, and disproportionate actions on the eye muscles in binocular movements, which is the simultaneous movement of the eyes. This test is commonly used in the detection, follow-up, and surgical planning of patients with strabismus. Strabismus is a condition in which the eyes do not have adequate alignment, generating several issues not only related to vision but also in social relations. This condition affects approximately 4% of the world’s population. In practice, the version exam is a subjective test, and the creation of a method that aims to automate the examination can help obtain a more objective result. Thus, this work presents an innovative computational method to perform the version examination automatically through face images. The proposed method is organized in seven main steps: (1) image acquisition; (2) preprocessing through a mean filter and Color Badger; (3) eye localization, based on skin segmentation and using Histograms of Oriented Gradients and Random Forest; (4) sclera segmentation, using statistical characteristics and the Random Forest classifier; (5) limbus localization, using as aid the segmented sclera; (6) eye corner localization, based on eye shape, analyzing the inner and outer eyes corners; and (7) version exam, using circle markers for each of the nine eye positions exams. The automatic measurement presented by the method was evaluated through the mean of the difference between the results provided by the method and the original versions measured by the specialist. When considering version, the proposed method obtained a mean accuracy and error of, respectively, 85.18% and 0.29 for the Medial Rectus muscle, 100% and 0 for the Lateral Rectus muscle, 86.9% and 0.47 for the Inferior Oblique muscle, 87.5% and 0.16 for the Superior Rectus muscle, 100% and 0 for the Superior Oblique muscle, and 95.23% and 0.28 for the Inferior Rectus muscle.

An Unsupervised Hybrid Symbolic Fuzzy Clustering Approach for Efficient Sclera Segmentation

This article performs the sclera segmentation task by proposing a new hybrid symbolic fuzzy c-means (HSFCM) clustering method. Practically, though the data point exhibits some sort of similarity, unfortunately they are not undistinguishable and exhibit some sort of dissimilarity. Thus, to capture these disparities, the proposed work uses symbolic interval valued representation method. Further, to handle uncertainty and imprecision, the paper has proposed to use symbolic fuzzy clustering methods. To assess the performance of the proposed method, extensive experimentation is conducted on SSRBC2016 dataset. The proposed clustering method is compared with existing FCM, KFCM, RSKFCM method in terms of cluster validity indices and accuracy. The obtained outcomes demonstrated that the proposed method performed better compared to the contemporary methods.

Sclera Recognition Based on Efficient Sclera Segmentation and Significant Vessel Matching

Abstract The visible blood vein in sclera has unique characteristics for each person, and it can be captured in visible light conditions. Therefore, the sclera vessels have been used as an important feature to improve the performance of identification recognition systems, for example, mobile payment and phone encryption system. But recent studies show that unsatisfactory sclera segmentation and tedious matching process are the two critical issues to degrade the performance of sclera identification system. In this paper, we propose a robust sclera recognition system with high accuracy and efficiency. First, a novel sclera segmentation method that provides adaptive threshold is proposed. Second, we have designed an improved vessels enhancement method based on the local grey distribution and local texture information, and least square linearization and moment invariants are utilized to extract features. Finally, an efficient matching strategy is put forward based on the detected significant vessels. The experimental results on one database with images captured in colour illumination prove that the proposed sclera recognition method has an obvious improvement regarding efficiency and accuracy over other sclera recognition systems.

A New Approach for Sclera Segmentation Using Integro Differential Operator

Sclera vessel patterns are one of the most reliable biometrics against performance, accessibility and availability. Sclera segmentation is the important phase of sclera recognition system as its performance depends on the segmentation accuracy. The big challenge in segmentation is nonlinearity of iris, sclera and upper and lower eyelid. This paper proposes a new approach to isolate the sclera accurately from the unconditionally acquired eye image. Hough transform and Integro differential operator (IDO) are popularly used to confine the iris and sclera boundary. Both of these two localizations are done based on the midpoint of the iris. These techniques produce better results as compared with other approaches but computation of iris center point incurs more time. To overcome this time consumption issue gabor based filter has been used. Experimental results are listed based on the UBIRIS database images. The experimental result shows that the proposed segmentation system will produce fast and accurate result.

A comprehensive investigation into sclera biometrics: a novel dataset and performance study

The area of ocular biometrics is among the most popular branches of biometric recognition technology. This area has long been dominated by iris recognition research, while other ocular modalities such as the periocular region or the vasculature of the sclera have received significantly less attention in the literature. Consequently, ocular modalities beyond the iris are not well studied and their characteristics are today still not as well understood. While recent needs for more secure authentication schemes have considerably increased the interest in competing ocular modalities, progress in these areas is still held back by the lack of publicly available datasets that would allow for more targeted research into specific ocular characteristics next to the iris. In this paper, we aim to bridge this gap for the case of sclera biometrics and introduce a novel dataset designed for research into ocular biometrics and most importantly for research into the vasculature of the sclera. Our dataset, called Sclera Blood Vessels, Periocular and Iris (SBVPI), is, to the best of our knowledge, the first publicly available dataset designed specifically with research in sclera biometrics in mind. The dataset contains high-quality RGB ocular images, captured in the visible spectrum, belonging to 55 subjects. Unlike competing datasets, it comes with manual markups of various eye regions, such as the iris, pupil, canthus or eyelashes and a detailed pixel-wise annotation of the complete sclera vasculature for a subset of the images. Additionally, the datasets ship with gender and age labels. The unique characteristics of the dataset allow us to study aspects of sclera biometrics technology that have not been studied before in the literature (e.g. vasculature segmentation techniques) as well as issues that are of key importance for practical recognition systems. Thus, next to the SBVPI dataset we also present in this paper a comprehensive investigation into sclera biometrics and the main covariates that affect the performance of sclera segmentation and recognition techniques, such as gender, age, gaze direction or image resolution. Our experiments not only demonstrate the usefulness of the newly introduced dataset, but also contribute to a better understanding of sclera biometrics in general.

ScleraSegNet: An Attention Assisted U-Net Model for Accurate Sclera Segmentation

This paper proposes a novel sclera segmentation approach based on an attention assisted U-Net model, named ScleraSegNet. Several off-the-shelf or improved attention modules are incorporated into the central bottleneck part or skip connection part of the original U-Net, helping the new model implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the necessity of using an external ROI localization model of cascade frameworks. The proposed approach is evaluated on several public, challenging eye datasets and experimental results show that introduced attention modules consistently improve the segmentation performance over the original U-Net across different datasets, and the best performing ScleraSegNet (CBAM) model achieves state-of-the-art segmentation performance. By utilizing the high stage’s semantic information to guide the selection of features from the low stage in channel-wise and spatial-wise, the further improved ScleraSegNet (SSBC) model ranked first in the Sclera Segmentation Benchmarking Competition 2019 (SSBC 2019), part of ICB 2019, with a Precision value of 92.88% and a Recall value of 90.34% on the MASD.v1 dataset, and a Precision value of 83.19% and a Recall value of 80.01% on the MSD dataset, respectively.

Validation of Computerized Quantification of Ocular Redness.

PurposeTo show feasibility of computerized techniques for ocular redness quantification in clinical studies, and to propose an automatic, objective method.MethodsSoftware for quantification of redness of the bulbar conjunctiva was developed. It provides an interface for manual and automatic sclera segmentation along with automated alignment of region of interest to enable estimation of changes in redness. The software also includes the redness scoring methods: (1) contrast-limited adaptive histogram equalization (CLAHE) in red-green-blue (RGB) color model, (2) product of saturation and hue in hue-saturation-value (HSV), and (3) average of angular sections in HSV. Our validation pipeline compares the scoring outcomes from the perspectives of segmentation reliability, segmentation precision, segmentation automation, and the choice of redness scoring methods.ResultsNinety-two photographs of eyes before and after provoked redness were evaluated. Redness in manually segmented images was significantly different within human observers (interobserver, P = 0.04) and two scoring sessions (intraobserver, P < 0.001). Automated segmentation showed the smallest variability, and can therefore be seen as a robust segmentation method. The RGB-based scoring method was less sensitive in redness assessment.ConclusionsComputation of ocular redness depends heavily on sclera segmentation. Manual segmentation appears to be subjective, resulting in systematic errors in intraobserver and interobserver settings. At the same time, automatic segmentation seems to be consistent. The scoring methods relying on HSV color space appeared to be more consistent.Translational RelevanceComputerized quantification of ocular redness holds great promise to objectify ocular redness in the standard clinical care and, in particular, in clinical trials.

Sclera-Net: Accurate Sclera Segmentation in Various Sensor Images Based on Residual Encoder and Decoder Network

Sclera segmentation is revealed to be of noteworthy importance for ocular biometrics. The paramount step for biometric recognition methods is the segmentation of the area of interest, i.e., the sclera in our case. The sclera segmentation process plays a pivotal part in retaining the accuracy of the sclera-based recognition schemes by restraining the errors. However, accurate sclera segmentation in the images from various sensors in a real environment is quite challenging due to the saturated and/or defocused vessel patterns and the vessel structure, which has complex nonlinear deformations due to the multilayered sclera. With the development of deep learning algorithms, studies that are based on the sclera segmentation using convolutional neural networks (CNNs) have achieved promising results for sclera recognition. However, previous CNN-based methods are based on the repeated subsampling stages of convolution strides, or spatial pooling leads to losing much of the finer image structure that significantly decreases overall performance in tasks, such as semantic segmentation. In this paper, we present Sclera-Net, a residual encoder and decoder network that exploits identity and non-identity mapping residual skip connections to take benefit of the high-frequency information from the prior layers of both encoder and decoder networks to determine the accurate sclera region as well as other ocular regions. In this way, the finer image structure that was being lost due to repeated subsampling during convolution and pooling can be reutilized using residual skip connections to enhance overall performance. Furthermore, the proposed Sclera-Net does not enhance the performance on the cost of increasing depth, complexity, or the number of parameters. We performed comprehensive experiments and obtained optimum performance not only on sclera datasets but also on the iris datasets. In particular, we achieved an equal error rate and mean F1-score of 0.0093 and 96.2421, respectively, on the challenging SBVPI database, which is the best-reported result to date.

Automated Vision-Based High Intraocular Pressure Detection Using Frontal Eye Images.

Glaucoma, the silent thief of vision, is mostly caused by the gradual increase of pressure in the eye which is known as intraocular pressure (IOP). An effective way to prevent the rise in eye pressure is by early detection. Prior computer vision-based work regarding IOP relies on fundus images of the optic nerves. This paper provides a novel vision-based framework to help in the initial IOP screening using only frontal eye images. The framework first introduces the utilization of a fully convolutional neural (FCN) network on frontal eye images for sclera and iris segmentation. Using these extracted areas, six features that include mean redness level of the sclera, red area percentage, Pupil/Iris diameter ratio, and three sclera contour features (distance, area, and angle) are computed. A database of images from the Princess Basma Hospital is used in this work, containing 400 facial images; 200 cases with normal IOP; and 200 cases with high IOP. Once the features are extracted, two classifiers (support vector machine and decision tree) are applied to obtain the status of the patients in terms of IOP (normal or high). The overall accuracy of the proposed framework is over 97.75% using the decision tree. The novelties and contributions of this work include introducing a fully convolutional network architecture for eye sclera segmentation, in addition to scientifically correlating the frontal eye view (image) with IOP by introducing new sclera contour features that have not been previously introduced in the literature from frontal eye images for IOP status determination.