Glaucoma is a blindness causing eye disease if not treated in time and caused by the increase in the cup-to-disc region (CDR). A novel method for extraction of the inner limiting membrane (ILM) and retinal pigment epithelium (RPE) layers from optical coherence tomography scans is proposed. A new colour channels mean based quality assessment step is applied to segment out ILM layer for different quality images. During RPE segmentation, a new ‘centroid based thresholding’ method is proposed to remove extended ILM regions. The method uses both the ILM layer and RPE breakpoints for a cup and disc calculation, respectively. A novel criterion for horizontal/flat cup diameter based on the average RPE break points is proposed. Based on calculated CDRs, the system classifies the subject as normal or glaucomatous. The average sensitivity, accuracy, and specificity of the proposed system are 87, 79 and 72%, respectively, on the Armed Forces Institute of Ophthalmology dataset when correlated with clinical annotations and CDR computer generated values. The proposed system has shown the higher correlated result of 92.59% sensitivity with the senior most ophthalmologists. It promotes the e-health field at the retinal level and employed as the decision support system by the young doctors.

Information hiding is particularly used for security applications to protect the secret message from an unauthorised person. Due to the tremendous development of the Internet and its usage, the issue of protection over the internet is increasing. Under such a condition, transforming the information from the transmitter to the receiver requires more security. Accordingly, in my previous research, an efficient medical image watermarking technique in E-healthcare application using a combination of compression and cryptography algorithm was proposed. The system only gives confidentiality and reliability. To overcome the problem, the authors propose a biometric-based on an efficient medical image watermarking in E-healthcare application, which produces a system for authentication, confidentiality, and reliability of the system. The proposed system utilises the fingerprint biometric for authentication, cryptography process for confidentiality, and reversible watermarking for the integrity. Basically, the proposed system consists of two stages such as (i) watermark embedding process and (ii) watermark extraction process. The experiments were carried out on the different medical images with electronic health record and the effectiveness of the proposed algorithm is analysed with the help of peak signal-to-noise ratio and normalised correlation.

Recently, tooth decay detection is considered as one of the emerging topics. Many diagnostic techniques have been successfully presented to diagnose the problems. However, the complexity in the tooth decaying diagnosis ascends when the environs are moderately difficult. Thus, this study introduces a novel caries detecting model for the accurate detection of tooth cavities. The model is divided into two phases: feature extraction and classification. Here, the feature extraction is based on multi-linear principal component analysis (MPCA), and the classification is processed using renowned neural network (NN) classifier. The NN classifier is trained using the adaptive dragonfly algorithm (ADA) algorithm. The proposed MPCA model Non-linear Programming with ADA (MNP-ADA) performance is compared with other existing methods and the performance of the approach is analysed in terms of measures such as accuracy, sensitivity, specificity, precision, false positive rate, false negative rate, negative predictive value, false discovery rate, F ₁-score, and Mathews correlation coefficient. The performance of the proposed model is analysed in terms of feature analysis and classifier analysis by comparing other models and proves the superiority of the developed caries detection model.

In the present scenario, retinal image processing is toiling hard to get an efficient algorithm for de-noising and segmenting the blood vessel confined inside the closed curvature boundary. On this ground, this study presents a hybrid active contour model with a novel preprocessing technique to segment the retinal blood vessel in different fundus images. Contour driven black top-hat transformation and phase-based binarisation method have been implemented to preserve the edge and corner details of the vessels. In the proposed work, gradient vector flow (GVF)-based snake and balloon method are combined to achieve better accuracy over different existing active contour models. In the earlier active contour models, the snake cannot enter inside the closed curvature resulting loss of tiny blood vessels. To circumvent this problem, an inflation term with GVF-based snake is incorporated together to achieve the new internal energy of snake for effective vessel segmentation. The evaluation parameters are calculated over four publically available databases: STARE, DRIVE, CHASE, and VAMPIRE. The proposed model outperforms its competitors by calculating a wide range of proven parameters to prove its robustness. The proposed method achieves an accuracy of 0.97 for DRIVE & CHASE and 0.96 for STARE & VAMPIRE datasets.

Pose detection of small targets in poor imaging conditions like heavy occlusion and low resolution is still an open and challenging task in computer vision. For instance, detection of students' poses in classrooms that are even indistinguishable to human eyes remains a rather difficult task. Motivated by the success of convolutional feature merging and locality preserving, the authors propose a pose detection framework combining merged region of interest (ROI) pooling and locality preserving learning. Unlike usual object detection algorithms which use general top-level convolutional features as inputs, their method uses a merged ROI pooling structure to merge semantic feature and high-resolution feature from the last two levels of convolutional feature maps, so that this merged feature is made more expressive than the single-level feature. In addition, the locality feature-preserving learning is used in the last fully-connected layer. Through locality preserving learning, features belonging to the same class would be forced to be closer in the feature space, which enables the model with stronger classification ability. Experimental results show that the proposed method outperforms the state-of-the-art methods.

In this study, the author proposes a total variational (TV) driven image restoration using discrete orthogonal Stockwell transform (DOST) when the noise (in the image) is an outcome of a Poisson process. Stockwell transform or S-transform (ST) is well known for its efficiency in resolving spatio-frequency components with high accuracy compared with many other transforms such as short-term Fourier transform, wavelet transform etc. This property of ST makes it more suitable for many image processing applications such as image restoration and image inpainting. By deriving the objective function and constraints of the optimisation problem (image restoration problem) based on the ST coefficients, the model becomes more robust in terms of preserving high resolution in the spatio-frequency domain. Images are modelled as an outcome of a Poisson process in many medical and telescopic imaging applications. The Poisson noise corruption is mainly due to the lack of a sufficient number of photons to reconstruct the data. In this study, corrupted images are restored due to the Poisson process (by which the data is formed) using the DOST under a non-local TV framework. The model is analysed and compared with the state-of-the-art Poisson noise removal methods using visual and statistical measures.

Turbid underwater environment poses great difficulties for the applications of vision technologies. One of the biggest challenges is the complicated noise distribution of the underwater images due to the serious scattering and absorption. To alleviate this problem, this work proposes a deep pixel-to-pixel networks model for underwater image enhancement by designing an encoding–decoding framework. It employs the convolution layers as encoding to filter the noise, while uses deconvolution layers as decoding to recover the missing details and refine the image pixel by pixel. Moreover, skip connection is introduced in the networks model in order to avoid low-level features losing while accelerating the training process. The model achieves the image enhancement in a self-adaptive data-driven way rather than considering the physical environment. Several comparison experiments are carried out with different datasets. Results show that it outperforms the state-of-the-art image restoration methods in underwater image defogging, denoising and colour enhancement.

In this study, the authors introduce new solutions and improvements to the multi-channel blind image deconvolution problem. More precisely, authors’ contributions are threefold: (i) At first, a simplified version of the existing cross-relation method for blind system identification is proposed; but most importantly, the authors incorporate into the channel estimation cost function a sparsity constraint to deal with the challenging issue of channel order overestimation errors; (ii) then, once the channel identification is achieved, a new image restoration method based on the stack decoding algorithm is introduced; and (iii) finally, a refining approach using an ‘all-at-once’ optimisation technique with an improved mixed norm regularisation is considered. The performance of the proposed approach was evaluated using several numerical simulations. Blind system identification and image restoration tasks were evaluated with respect to several criteria: numerical complexity, robustness to noise effects and channel order estimation. The results obtained are promising and highlight the effectiveness of the proposed approach.

An improved single image fast deblurring algorithm based on hyper-Laplacian constraint is proposed. The algorithm is improved in three aspects: image blur kernel estimation sub-region selection, blur kernel precise estimation, and fast non-blind deconvolution. First, image amplitude and gradient are used as the basis of blur kernel estimation. On the basis of analysing the edge amplitude and gradient of the image, the image sub-region for blur kernel estimation is selected. Then the sparsity of the blur kernel is restricted by hyper-Laplacian, and the fast solving mode of alternately solving different variables is designed. The blur kernel information is accurately estimated. In the fast non-blind deconvolution restoration phase of the image, the regularised constraint term of the hyper-Laplacian model is improved and the image gradient distribution is constrained. The blind growth trend of the regional gradient near the strong edge can be suppressed well, and the deblurred image with clear edge structure is generated. Experimental results show that the proposed algorithm can achieve better image deblurring effect and high efficiency.

In certain applications, classification models have to be trained with small datasets. This study proposes a new deep neural network with a feature generalisation layer (FGL). First, instead of using a generative network for data augmentation, the FGL is modelled using a latent variable model to diversify features directly by sharing other layers. Then, dual-objective functions are defined to optimise the parameters of the network: one minimises the generation error and the other minimises the classification error. Finally, a parallel multibranch structure is used in the FGL to improve the convergence of model training. The classification accuracy obtained using various quantities of training samples increased up to 4.63% on the MNIST dataset, up to 3.00% on the CIFAR10 nature image dataset, over the reference model. These experimental results illustrate the effectiveness of the authors’ method for training classification models with small datasets.

To extend the application of the cognitive associative network and improve the performance of the target tracking method, a novel tracking method based on the new network is proposed, i.e. the cognitive associative network is transformed and used to perform the target tracking. The local hue histograms are used to model the target, and the circle matching criterion is used to locate the target, which can restrain the disturbance from the adjacent regions in the background. It can also adjust adaptively the size of the tracking target according to the local matching results. Therefore, the tracking method based on the cognitive associative network has good invariability to the scale, the rotation, and the partial occlusion. Simulation results show that the tracking method can perform the target tracking in the disturbance environment or in the scene of the complex motion. The tracking method can locate the target more accurately than the common tracking methods such as the Camshift method or the compressive tracking method.

Segmentation of lumen (LB) and outer wall boundaries (OB) of carotid artery in magnetic resonance (MR) images is essential for carotid atherosclerotic disease diagnosis. However, the limited image signal-to-noise ratio, flow artefact, and varied lumen and outer wall become significant obstacles for automatic segmentation. A fully automatic framework is proposed for LB and OB segmentation in MR images. First, the lumen is identified by the support vector machine using a special strategy and LB is segmented by the geodesic star-shape-constrained graph cut. Then a novel global optimisation is developed to segment OB based on the graph cut, which consists of shape priors and appearance priors. The shape priors are learned from labelled shapes on LB and OB, while the appearance priors are modelled by Gaussian mixture models. A novel shape constraint is also designed as the constraint term. To evaluate author's method, extensive experiments are carried out from 160 MR images belonging to 16 patients. Experimental results demonstrate that the proposed method can yield high accuracy with fully automatic segmentation. Moreover, the advantages of the proposed method have been shown in terms of high flexibility and accuracy without user interactions in comparison with other methods.

This work addresses the problem of searching and retrieving similar textual images based on the detected text and opens the new directions for textual image retrieval. For image retrieval, several methods have been proposed to extract visual features and social tags; however, to extract embedded and scene text within images and use that text as automatic keywords/tags is still a young research field for text-based and content-based image retrieval applications. The automatic text detection retrieval is an emerging technology for robotics and artificial intelligence. In this study, the authors have proposed a novel approach to detect the text in an image and exploit it as keywords and tags for automatic text-based image retrieval. First, text regions are detected using maximally stable extremal region algorithm. Second, unwanted false positive text regions are eliminated based on geometric properties and stroke width transform. Next, the true text regions are proceeded into optical character recognition for recognition. Third, keywords are formed using a neural probabilistic language model. Finally, the textual images are indexed and retrieved based on the detected keywords. The experimental results on two benchmark datasets show the dominancy of text is efficient and valuable for image retrieval specifically for textual images.

Inter-frame forgery marks a central type of forgery in surveillance videos, and involves three aspects – frame duplication, insertion, and deletion – under temporal domain. However, this forgery type has received little attention from scholars. More efforts have been on detecting only a single aspect of inter-frame forgery. Furthermore, studies have confirmed that previous methods did not achieve high accuracy for all forgeries types with low computational loads at the same time. In this study, the proposed method establishes a framework that can simultaneously detect all aspects of inter-frame forgeries. During the decoding process, the authors extract residue data of each frame from a video stream. Then spatial and temporal energies are exploited to illustrate data flow, and abnormal points are determined to detect forged frames. Noise ratios of forged and original frames are estimated for differentiating insertion from duplication attacks. Experimental results indicate that the proposed method achieves higher accuracy and lower computational time for detecting inter-frame forgery.

Telemedicine systems require reliable, small size, and secure transmission. In teledermoscopy, the transmission of the traditional large size dermoscopic images leads to heavy consumption of storage space and bandwidth congestion, which impairs the efficient use of the transmission channel bandwidth. Furthermore, halftoning and inverse halftoning have a great practical impact in the reduction of the images’ size. For teledermoscopy, the present work optimises the error filter during the error diffusion process to generate the halftone dermoscopic images with small size. The proposed inverse halftone optimises two parameters during the integration between the regularised inverse filter and the DLPA-ICI (directional local polynomial approximation and the intersection of confidence intervals) estimator's kernel. The optimised kernel parameter determines the best window size of the kernel in the restored image estimation. The experimental results included comparative studies in terms of several image quality metrics and the average size reduction in percentage when using the genetic algorithm for optimisation with Jarvis error diffusion for halftone images generation. The achieved average size reduction is 92.15% using the proposed method compared to the average dermoscopic images’ size in the test set on 900 images showing the superiority of the proposed method.

Low-light enhancement methods suffer from the over-enhancement problem which could induce the loss of the important texture and make images look unnatural. Moreover, some low-light images contain strong light areas that must be weakened to improve the visual effect. In this study, an enhancement method with strong light weakening and bright halo suppressing is presented. Firstly, the bright channel prior is applied to the inverted image to weaken the strong light both in and around the strong light areas. Then, a dehazing-type of algorithm with the dark channel prior is employed via superpixel segmentation to enhance the low-light image. Finally, a revised non-local denoising method is proposed to further refine the enhanced image. Experimental results showed that the proposed method achieved better visual effects compared with other state-of-the-art methods. Besides, the quantitative evaluation showed that the authors’ method outperforms the other methods both in the aspect of enhancement and denoising.

Lung cancer is one of the deadliest diseases worldwide and the classification of different types of lung cancers in computed tomography (CT) images is also one of the most significant issues in computer-aided diagnosis. It remains a tough task since various image features could be extracted from one single image while part of the features is irrelevant to the final diagnosis results. In this study, a knockoff filter-based approach is proposed to produce the optimal feature set and to minimise the irrelevancy of the output features for the classification of lung cancer in CT images. The proposed feature selection strategy not only can generate the optimal feature subset but also constrain the false discovery rate of the irrelevant features under a specified parameter setting. Ten-fold leave-one-out cross-validation and the area under the receiver operating characteristic curve are both adopted in the experiments to evaluate the performance of the proposed method. The areas under curve of is achieved when the support vector machine classifier is trained on the features determined by the proposed feature selection strategy. The experimental results demonstrate that the presented approach is potentially valuable for lung cancer diagnosis.