Milano-Retinex is a family of Retinex-inspired spatial colour algorithms mainly developed for colour image enhancement. According to the Retinex theory, a Milano-Retinex algorithm takes as input an RGB image and processes the colour intensities of each pixel (i.e. the target) based on the spatial distribution of the colour intensities sampled in a surrounding region. The output is an RGB image, with locally adjusted colours and contrast. In Milano-Retinex family, different ways of spatial sampling are implemented. This study reviews and compares these sampling characteristics within a group of Milano-Retinex algorithms developed in the last decade, from Random Spray Retinex (2007) to the gradient-based colour sampling schemes GREAT and GRASS (2017). Instead of exploring the target neighbourhood by random paths as the original Retinex algorithm does, these methods consider sets of pixels, randomly or deterministically defined, including all the image pixels or a part of them, such as random sprays or image edges. They replace the ratio-reset-threshold-product-average mechanism of the original Retinex with equations re-working maximal intensities over the sampled sets. The performance of these approaches is compared with more than 200 images of indoor and outdoor scenes, captured by commercial cameras under several different conditions.

Similarity searching of high-dimensional data is fundamental in the multimedia research field. In recent years, the binary code indexing has achieved significant applications in the context of similarity searching. However, most of the existing binary coding methods adopt a random generation method in near neighbour cluster problems, which involve unnecessary computations and degrade similarity in object points. To avoid the uncertainty of random generation codes, in this study, the authors propose a new locality sensitive hashing (LSH) algorithm based on q -ary Bose–Chaudhuri–Hocquenghem (BCH) code. BCH–LSH algorithm utilises the characteristics of the designed distance of BCH codes and uses the BCH codes generator matrix as a transform basis of the hash function to map the source data into the hash space. The experiments show that the BCH–LSH algorithm is superior to the E2LSH algorithm in average precision, average recall ratio and running speed.

Star images obtained by star sensors have a low signal-to-noise ratio due to various physical constraints. Low resolution also causes stellar centroid extraction error when traditional methods such as the Gaussian filter or adaptive median filter are utilised to de-noise star images. An automatic centroid extraction method for noisy low-resolution star images is proposed in this study. First, sparse representation is utilised to de-noise the Poisson–Gaussian mixed noise of the low-resolution star image. A high-resolution star image is then reconstructed by using the low-resolution sparse coefficient. Finally, the stellar centroids are extracted automatically by learning the relationship between the high-resolution star image and corresponding stellar centroid image. Experimental results indicate that the positioning accuracy of the stellar centroids is also greatly enhanced by the reconstructed high-resolution stellar centroid image. The correct rate of stellar centroid recognition is 99.35%; the positioning accuracy of stellar centroid and computing time are 16.21″ and 11.30 ms, respectively. The probability distributions ofPoisson and Gaussian noises are 0.50 and 0.08, respectively, while the proposed method correctly recognises stellar centroids at a rate of 76.56%. The results presented here may provide a workable foundation for accurate attitude calculations of the celestial navigation system.

To overcome the drawbacks of existing filters for salt and pepper noises, an adaptive probability filter is proposed. For an image, it detects salt and pepper noises based on the characteristic of minimum and maximum intensity values of the images, as well as the distribution of noise. If the noise-free intensities in neighbourhood repeat with a certain probability, the noise-free intensity with highest repeated frequency is used to remove noise based on the statistical significance; otherwise, the median of noise-free pixels in neighbourhood is used to remove noise. Experiments show that the proposed method is capable of detecting noise more accurately and perform much better than the existing distinguished filters in terms of peak-signal-to-noise ratio, image enhancement factor, and visual representation at all the noise densities.

Currently, the structural similarity index metric (SSIM) is recognised generally and applied widely in image quality assessment (IQA). However, using SSIM to evaluate contrast-distorted images from TID2013 and CSIQ databases is low effective. In this study, the authors improve SSIM for contrast-distorted images by combining it with the contrast sensitivity characteristics of human visual system (HVS). In the improved method, first, they combine the visual characteristics to propose a model that HVS perceives the real image. Then, this model is used to eliminate the visual redundancy of real images. Afterwards, the perceived images are evaluated using SSIM. Furthermore, 241 contrast-distorted images from TID2013 and CSIQ databases were used in experiments. The results have shown that in comparison with SSIM scores, the scores obtained by the improved SSIM are more consistent with the subjective assessment scores. Moreover, the Pearson linear correlation coefficient and Spearman rank order correlation coefficient between subjective and objective scores are averagely improved by 12.83 and 22.78%, respectively. In addition, the assessment accuracy of the improved SSIM is better than that of five commonly used IQA metrics. Also, it has an excellent generalisation performance. These results show that the assessment performance of the improved SSIM is effectively enhanced.

Image fidelity refers to the ability of a process to render an image accurately. As image enhancement algorithms have been developed in recent years, how to assess the performances of different image enhancement algorithms has become an important question. Some objective image quality assessment (IQA) methods have been proposed, but there is little research on the image fidelity evaluation when comparing the performances of enhancement algorithms. Therefore, the authors proposed a new image fidelity assessment framework consisting of three components: the information entropy fidelity, constituent fidelity and colour fidelity. To verify the rationality of the fidelity criteria, they used the popular IQA database (LIVE), and the results indicated that the method matched better with the subjective assessment. Then, they verified the effectiveness of their method with the famous technique of image enhancement: multi-scale Retinex with colour restoration (MSRCR). The experimental results demonstrate MSRCR can improve the image quality, but it gives rise to obvious distortions. It is necessary to keep a moderate balance between image fidelity and image quality when they assess the enhanced images. Their results showed that the proposed objective fidelity index could provide an additional objective basis for the quality evaluation of image enhancement algorithms.

Small geometric misalignments of micro computed tomography (CT) system will cause geometric artefacts in the reconstructed image. A new correction method of geometrical artefacts based on marker and non-linear optimisation model is proposed. In this method, the simple balls marker and the measured objects are scanned simultaneously, and the geometric parameters of the micro-CT system are precisely estimated by solving the non-linear optimisation model which is based on the scanning data. The geometric artefacts caused by geometric parameters are corrected and the authors can reconstruct the image correctly. In addition to estimating geometric parameters for the traditional scanning mode, the proposed method can also be used for the limited angle CT scanning and the half detector CT scanning. Simulated experiments and real experiments verify that the correction method effectively decrease the geometric artefacts of micro-CT images.

Recognising human activity from video stream has become one of the most interesting applications in computer vision. In this study, a novel hybrid technique for human action recognition is proposed based on fast HOG3D of integral videos and Smith–Waterman partial shape matching of the fused frame. The proposed technique is divided into two main stages, the first stage extracts a set of foreground snippets from the input video, and extracts the histogram of 3D gradient orientations from the spatio-temporal volumetric data; and the second stage fuses a set of key frames from current snippet and extracts the contours from the fused frame. Non-linear support vector machine (SVM) decision trees are used to classify HOG3D features into one of fixed action categories. On the other hand, Smith–Waterman partial shape matching is used to compare between the contour of the fused frame and the stored template contour of specified action. The results from SVM and Smith–Waterman partial shape matching are then combined. The experimental results show that combining non-linear SVM decision trees of HOG3D features and Smith–Waterman partial shape matching of fused contours improved the accuracy of the classification model while maintaining efficiency in time elapsed for training.

This study presents an image denoising technique using multiscale non-local means (NLM) filtering combined with hard thresholding in curvelet domain. The inevitable ringing artefacts in the reconstructed image – due to thresholding – is further processed using a guided image filter for better preservation of local structures like edges, textures and small details. The authors decomposed the image into three different curvelet scales including the approximation and the fine scale. The low-frequency noise in the approximation sub-band and the edges with small textural details in the fine scale are processed independently using a multiscale NLM filter. On the other hand, the hard thresholding in the remaining coarser scale is applied to separate the signal and the noise subspace. Experimental results on both greyscale and colour images indicate that the proposed approach is competitive at lower noise strength with respect to peak signal to noise ratio and structural similarity index measure and excels in performance at higher noise strength compared with several state-of-the-art algorithms.

Age-related macular degeneration (AMD) is one of the main reasons for visual impairment worldwide. The assessment of risk for the development of AMD requires reliable detection and quantitative mapping of retinal abnormalities that are considered as precursors of the disease. Typical signs of the latter are the so-called drusen that appear as yellowish spots in the retina. Automated detection and segmentation of drusen provide vital information about the severity of the disease. The authors propose a novel method for the detection and segmentation of drusen in colour fundus images. The method combines colour information of the object with its boundary information for the accurate detection and segmentation of drusen. To perform non-uniform illumination correction and to minimise inter-subject variability a novel colour normalisation method has been proposed. Experiments are conducted on publicly available STARE and ARIA datasets. The method achieves an overall accuracy of 96.62% which is about 4% higher than the state-of-the-art method. The sensitivity and specificity of the proposed method are 95.96 and 97.64%, respectively.

Here, the authors propose a spatially weighted super-resolution (SR) algorithm, which takes into consideration the distribution of every information that characterise different image areas. The authors investigate to use a combined spatially weighted regularisation of the bilateral total variation and a second-order term increasing then the robustness of the proposed SR approach with respect to blur and noise degradations. In addition, the authors propose an iterative Bregman iteration algorithm to resolve the obtained optimisation SR problem. As a result, this regularisation is more efficient and easier to implement; moreover, it preserves well the smooth regions of the image and also sharp edges. Using different simulated and real tests, the authors prove the efficiency of the proposed algorithm compared to some SR methods.

An improved version of supervised locally linear embedding is proposed. In this algorithm, the weight factors of the supervised method are adaptively achieved. This method can simplify the supervised feature extraction algorithm by reducing parameters. To improve classification accuracy, a clustering-based fuzzy support vector machine (FSVM) is proposed. Different from traditional FSVMs, the proposed method constructs the fuzzy weights by inner-class clusters. In the proposed method, loose density is defined to express the compactness of the inner-class clusters. The proposed algorithm can restrain the noise and outliers by exploiting the method of endowing with smaller weight for big loose density and bigger weight for the small loose density of samples in the clusters. To inspect the performance of the proposed methods, we conduct experiments on two hyper-spectral images. Results show that the two methods are competitive among the competitors.

In this study, the authors describe a modified non-blind and blind deconvolution model by introducing a regularisation parameter that incorporates the spatial image information. Indeed, they have used a weighted total variation term, where the weight is a spatially adaptive parameter based on the image gradient. The proposed models are solved by the split Bregman method. To handle adequately the discrete convolution transform in a moderate time, fast Fourier transform is used. Tests are conducted on several images, and for assessing the results, they define appropriate weighted versions of two standard image quality metrics. These new weighted metrics clearly highlight the advantage of the spatially adaptive approach.

Matrix completion is to recover a low-rank matrix from a subset of its entries. One of the solution strategies is based on nuclear norm minimisation. However, since the nuclear norm is defined as the sum of all singular values, each of which is treated equally, the rank function may not be well approximated in practice. To overcome this drawback, this study presents a matrix completion method based on capped nuclear norm (MC-CNN). The capped nuclear norm can reflect the rank function more directly and accurately than the nuclear norm, Schatten p-norm (to the power p) and truncated nuclear norm. The relation between the capped nuclear norm and the truncated nuclear norm is revealed for the first time. Difference of convex functions’ programming is employed to solve MC-CNN. In the proposed algorithm, a key sub-problem, i.e. a matrix completion problem with linear regularisation term, is solved by using the active subspace selection method. In addition, the algorithm convergence is discussed. Experimental results show encouraging results of the proposed algorithm in comparison with the state-of-the-art matrix completion methods on both synthetic and real visual datasets.

The authors propose efficient hybrid error resilience and error concealment (ER-EC) algorithms for H.264 3D video-plus-depth (3DV + D) transmission over error-prone channels. At the encoder, content-adaptive pre-processing ER mechanisms are implemented by applying the context adaptive variable length coding (CAVLC), the slice structured coding, and the explicit flexible macro-block ordering. At the decoder, a post-processing EC algorithm with multi-proposition schemes is implemented to recover the lost 3DV colour frames. The convenient EC hypothesis is adopted based on the lost macro-blocks size mode, the faulty view, and the frame types. For the recovery of the lost 3DV depth frames, an encoder-independent decoder-dependent depth-assisted EC algorithm is suggested. It exploits the previously estimated colour disparity vectors (DVs) and motion vectors (MVs) to estimate more additional depth-assisted MVs and DVs. After that, the optimum colour-plus-depth DVs and MVs are accurately selected by employing the directional interpolation EC algorithm and the decoder MV estimation algorithm. Finally, a weighted overlapping block motion and disparity compensation scheme is utilised to reinforce the performance of the proposed hybrid ER-EC algorithms. Experimental results on standard 3DV + D sequences show that the proposed hybrid algorithms have superior objective and subjective performance indices.

This study presents a hierarchical palmprint feature extraction and recognition approach based on multi-wavelet and complex network (CN) since they can effectively decrease redundant information and enhance key points of main lines and wrinkles. The palmprint is first pre-filtered and decomposed once using multi-wavelet. Three components (LL_1,2,3) corresponding to the pre-filter except for diagonal component are extracted as the elementary features. Second, binary images (BLL_1,2,3) are obtained by the average window method using different thresholds. Third, three series of dynamic evolution CN models (the 1st, 2nd, 3rd CNs) are constructed from global to local, which is based on the mosaiced images obtained from BLL_1,2,3, BLL₁ and four equally divided sub-images of BLL₁, respectively. Fourth, statistical features are extracted from complex networks, in which average degree and standard deviation of the degrees are extracted for the 1st CNs and average degrees are extracted for the 2nd and 3rd CNs. Fifth, the fisher feature is extracted using the linear discriminate analysis method. Finally, the nearest neighbourhood classifier is used to recognise palmprint. Based on the CASIA Palmprint Image Database, experimental results show that the proposed method can effectively recognise palmprint with good robustness and overcome the problem of small training samples number.

Psychophysical and neurophysiological investigations on the human visual system show that most neurons in the primary visual cortex (V1) possess a non-classical receptive field (nCRF) region in addition to the CRF region. The nCRF has a modulatory, normally inhibitory, effect on the responses to visual stimuli generated within the CRF. In computational terms, this mechanism suppresses the response to edges in the presence of similar edges in the surroundings. Many computational techniques have been proposed to address the surround suppression mechanism. These methods introduce an inhibition term that is required to suppress the textures and protect the contours. Several studies have found that the spatial summation properties over the receptive fields of retinal X cells are approximately linear, while they are non-linear for Y cells. Inspired by the visual information processing in the X–Y channel and spatial summation properties of X and Y cells, the authors propose a contour detector using linear and non-linear modulations based on nCRF suppression. Extensive experimental evaluations demonstrate that their contour detector significantly outperforms other algorithms. The methods proposed in this study are expected to facilitate the development of efficient computational models in the field of machine vision.

Here, a hybrid feature descriptor-based method is proposed to recognise human emotions from their facial expressions. A combination of spatial bag of features (SBoFs) with spatial scale-invariant feature transform (SBoF-SSIFT), and SBoFs with spatial speeded up robust transform are utilised to improve the ability to recognise facial expressions. For classification of emotions, K-nearest neighbour and support vector machines (SVMs) with linear, polynomial, and radial basis function kernels are applied. SBoFs descriptor generates a fixed length feature vector for all sample images irrespective of their size. Spatial SIFT and SURF features are independent of scaling, rotation, translation, projective transforms, and partly to illumination changes. A modified form of bag of features (BoFs) is employed by involving feature's spatial information for facial emotion recognition. The proposed method differs from conventional methods that are used for simple object categorisation without using spatial information. Experiments have been performed on extended Cohn–Kanade (CK+) and Japanese female facial expression (JAFFE) data sets. SBoF-SSIFT with SVM resulted in a recognition accuracy of 98.5% on CK+ and 98.3% on JAFFE data set. Images are resized through selective pre-processing, thereby retaining only the information of interest and reducing computation time.

The aim of the pansharpening scheme is to improve the spatial information of multispectral images using the panchromatic (PAN) image. In this study, a novel pansharpening scheme based on two-dimensional discrete fractional Fourier transform (2D-DFRFT) is proposed. In the proposed scheme, PAN and intensity images are transformed using 2D-DFRFT and filtered by highpass filters, respectively. The filtered images are inverse transformed and further used to generate the pansharpened image using appropriate fusion rule. The additional degree of freedom in terms of its angle parameters associated with the 2D-DFRFT is exploited for obtaining better results in the proposed pansharpening scheme. Simulation results of the proposed technique carried out in MATLAB are presented for IKONOS and GeoEye-1 satellite images and compared with existing fusion methods in terms of both visual observation and quality metrics. It is seen that the proposed pansharpening scheme has improved spectral and spatial resolution as compared to the existing schemes.

Here, the authors present a hardware design of fast multiplierless forward binary discrete cosine transform (BinDCT) based on quantum-dot cellular automata (QCA) technology. This new technology offers several features such as: small size, ultralow power consumption, and can operate at 1 THz. The simulation results in QCA Designer software confirm that the proposed circuit works well and can be used as a high-performance design in QCA technology. The analysis obtained from the implementation of QCA BinDCT indicates that the proposed architecture is superior to the existing based on classic metal-oxide (complementary metal-oxide semiconductor technology) technology. Here, the authors are going to introduce highly BinDCT module scaled with ultra-low power consuming. The proposed circuit requires 50% fewer power consuming compared to previous existing designs. The proposed architecture can attain a throughput of 800 mega pixel per second (Mpps). To design and verify the proposed architecture, QCADesigner tool and QCAPro tool are, respectively, employed for synthesis and power consumption estimation. Since the works in the field of QCA logic image processing have only started to bloom, the proposed contribution will engender a new thread of research in the field of real-time image and video treatment.

Images contain significant amounts of information but present different challenges relative to textual information. One such challenge is compound figures or images made up of two or more subfigures. A deep learning model is proposed for compound figure detection (CFD) in the biomedical article domain. First, pre-trained convolutional neural networks (CNNs) are selected for transfer learning to take advantage of the image classification performance of CNNs and to overcome the limited dataset of the CFD problem. Next, the pre-trained CNNs are fine-tuned on the training data with early-stopping to avoid overfitting. Alternatively, layer activations of the pre-trained CNNs are extracted and used as input features to a support vector machine classifier. Finally, individual model outputs are combined with score-based fusion. The proposed combined model obtained a best test accuracy of 90.03 and 96.93% outperforming traditional hand-crafted and other deep learning representations on the ImageCLEF 2015 and 2016 CFD subtask datasets, respectively, by using AlexNet, VGG-16, VGG-19 pre-trained CNNs fine-tuned until best validation accuracy stops increasing combined with the combPROD score-based fusion operator.

Development of automatic disease detection and classification system is significantly explored in precision agriculture. In the past few decades, researchers have studied several cultures exploiting different parts of a plant. A similar study is performed for Soybean using leaf images. A rule based semi-automatic system using concepts of k-means is designed and implemented to distinguish healthy leaves from diseased leaves. In addition, a diseased leaf is classified into one of the three categories (downy mildew, frog eye, and Septoria leaf blight). Experiments are performed by separately utilising colour features, texture features, and their combinations to train three models based on support vector machine classifier. Results are generated using thousands of images collected from PlantVillage dataset. Acceptable average accuracy values are reported for all the considered combinations which are also found to be better than existing ones. This study also attempts to discover the best performing feature set for leaf disease detection in Soybean. The system is shown to efficiently compute the disease severity as well. Visual examination of leaf samples further proves the suitability of the proposed system for detection, classification, and severity calculation.

Images degraded by haze usually have low contrast and fide colours, and thus have bad effects on applications such as object tracking, face recognition, and intelligent surveillance. So the purpose of dehazing is to recover the image contrast without colour distortion. The dark channel prior (DCP) is widely used in the field of haze removal because of its simplicity and effectiveness. However, when faced with bright white objects, DCP overestimates the haze from its true value and thus causes colour distortion. In this study, the authors propose a dehazing model combining saliency detection with DCP to obtain recovered images with little colour distortion. There are three main contributions. First, they introduce a novel saliency detection method, focusing on superpixel intensity contrast, to extract bright white objects in the hazy image. Those objects are not used to estimate the atmospheric light and transmission in the dark channel image. Second, a self-adaptive upper bound is set for the scene radiance to prevent some regions being too bright. Third, they propose a quantitative indicator, colour variance distance, to evaluate the colour restoration. Experimental results show that their proposed model generates less colour distortion and has better comprehensive performance than competing models.

In recent years, biometric applications have significantly gained popularity. Such applications involve voluminous databases of high dimensional data. These enormous databases increase the cost of identification and degrade the system performance. To resolve such an issue a plethora of algorithms based on geometric hashing, k–d tree, k-means clustering, etc., have been proposed in the literature. Although, these algorithms solve a number of concomitant challenges of multi-dimensional data, yet, they fail to present a universal solution. In this study, we propose an indexing mechanism, which partitions the data space effectively into zones and blocks using a set of hash functions. Furthermore, the index locations are divided into maximum nine sub-locations to store data. This helps in carrying out an efficient search of the queried data, thereby minimising the false acceptance and rejection rate. To validate the proposed approach, the mechanism has been applied to the fingerprint verification competition and National Institute of Standards and Technology fingerprint image databases. The experimental results substantiate the efficacy of our approach in terms of accuracy, speed, reduction of search space and the number of comparisons to store and retrieve data.

Steganography is used for secret or covert communication. A graph wavelet transform-based steganography using graph signal processing (GSP) is presented, which results in better visual quality stego image as well as extracted secret image. In the proposed scheme, graph wavelet transforms of both the cover image and transformed secret image (using Arnold cat map) are taken followed by alpha blending operation. The GSP-based inverse wavelet transform is performed on the resulting image, to get the stego image. Here, the use of GSP increases the inter-pixel correlation that results in better visual quality stego and extracted secret image as shown in simulation results. Simulation results show that the proposed scheme is more robust than other existing steganography techniques.