In this study, the authors propose a wavelet domain still image watermark detection method which uses the Bessel K probability density function to describe the distribution of wavelet coefficients. In this study, watermark detection is formulated as a binary statistical decision problem which is to detect a signal submerged in the noise that follows a Bessel K distribution. Using this formulation, an optimal watermark detector using likelihood ratio test is proposed. The experimental results of the proposed method in a variety of situations demonstrate that the proposed method has a robust detection performance for additive spread spectrum watermarks.

In this study, an efficient image fusion framework for multi-focus images is proposed based on compressed sensing. The new fusion framework consists of three parts: image sampling, measurement fusion and image reconstruction. First, the dual-channel pulse coupled neural network model is used in the image sampling part as an important weighting factor in the fusion scheme. Second, the result from the measurement fusion part is reconstructed through a new reconstruction algorithm called self-adaptively modified Landwebber filter. Finally, computer simulation-based experiment is conducted, showing that the novel fusion framework is capable of saving computational resource and enhancing the fusion result and is easy to implement.

Colour fringe is an artefact that mainly appears at the boundary of an object in an image, where colours change abruptly and luminance value is large. Thus, colour fringe should be corrected in a colour digital image. This study proposes a method for detection and correction of colour fringe in YC _b C _r colour space. The luminance value as well as gradient magnitudes/directions of YC _b C _r are used to detect colour fringed region. Then, chrominance values in colour fringed regions are corrected using those of pixels in the neighbouring regions along the colour fringe direction. The experimental results show the effectiveness of the proposed colour fringe detection and correction method.

In this study, the authors propose a novel method for removing mixed (mixture of impulse and Gaussian) multi-channel noise from multi-channel digital images based on a modified version of the algorithm introduced by Struyf and Rousseeuw (Comput. Stat. Data Anal. (2000), 34, pp. 415–426) for finding approximate halfspace deepest location (Tukey's median). Denoising results of this new nonlinear spatial domain filtering method applied to benchmark images corrupted by multi-channel mixed noise outperform currently used spatial domain filters and state-of-the-art wavelet transform domain filters in terms of both peak signal-to-noise ratio and visual quality. Unlike most of the existing algorithms which remove the noise from multi-channel digital images on each of the channels separately, our method, because of its multivariate/multi-dimensional nature, eliminates the noise on all channels simultaneously without their separation, thus preserving the spectral correlation between channels in a multi-channel image. Proposed denoising method is very effective for removal of very wide range of powers of mixed multi-channel noise, but can be also successfully implemented for reduction of other forms of multi-channel noise since it is independent of the source or distribution of the noise.

As digital sports video becomes increasingly pervasive, semantic video summary becomes one of the important components for the next generation of multimedia applications. Ontology is a feasible way to mine the semantic information from the video stream. However, current ontology-based methods did not concentrate on the effectiveness and soundness of semantic reasoning. Here, the authors propose a content-directed ontology reasoning approach to produce meaningful sports video summarisation. The proposed ontology can facilitate the metadata acquisition of video and the improvement of query performance. It also provides a flexible way to query the sports video database, which cannot be achieved by simple keyword search. For annotating, describing and managing the sports video content, we propose a sports video descriptive language (SVDL) based on the proposed ontology. Moreover, the semantically meaningful sports video abstraction is produced by reasoning engine which is based on the extension of the Tableau algorithm. Meanwhile, the soundness and completeness of the reasoning algorithm can be solidly proved. Subjective assessment experimental results reveal the reliability and efficiency of the propose scheme.

Fourth-order partial differential equations (PDEs) for noise removal are able to provide a good trade-off between noise removal and edge preservation, and can avoid blocky effects often caused by second-order PDE. In this study, the authors propose a fourth-order telegraph-diffusion equation (TDE) for noise removal. In the authors method, a domain-based fourth-order PDE is proposed, which takes advantage of statistic characteristics of isolated speckles in the Laplace domain to segment the image domain into two domains: speckle domain and non-speckle domain. Then, depending on the domain type, they adopt different conductance coefficients in the proposed fourth-order PDE. The proposed method inherits the advantage of fourth-order PDE which is able to avoid the blocky effects widely seen in images processed by second-order PDE. Furthermore, a TDE processing scheme is derived from previously proposed domain-based fourth-order PDE by adding second time derivative, which results in better edge preservation, whereas yielding better improvement in signal-to-noise ratio and low noise sensitivity. Experimental results show the effectiveness of the proposed method.

Registration between catadioptric omni-images and orthorectified aerial images is the key step to integrate them to achieve three-dimensional urban construction. This problem becomes very challenging because of the non-linearity of the imaging model of catadioptric omni-cameras. In this study, the authors attempt to address this problem. The authors first study the properties of horizontal line structure under catadioptric omni-cameras to prove and extend the theorem of catadioptric distance, and then present angle consistency of horizontal lines between a catadioptric omni-image and an orthorectified aerial image. The authors further employ them to achieve registration between catadioptric omni-images and orthorectified aerial images. To the best of authors’ knowledge, this study has not been done before. Experimental results on both simulated data and real scene images confirm the effectiveness of this approach.

In this study a set of colour gradients based on colour visual perception, which use International Commission on Illumination (CIE) L*a*b* colour space, is presented. The main objective is the study of how the colour difference equations, developed by CIE, affect the estimation of the gradients in terms of correlation with colour visual perception. To evaluate the gradients performance they are used as the basis of an edge detector based on levelset. A set of synthetic images was designed to evaluate which edge detector and consequently, which colour difference equation, is more correlated with human perception of colour. Both quantitative and qualitative measurements showed that the results obtained using CIE94 have a higher correlation with what the human eye can perceive.

This study proposes a new method for the detection of local invariant features with contour. This method differs from traditional methods that use image intensity. Image contours can be extracted stably with changes in viewpoint, scale, illumination and other factors. The proposed algorithm first extracts the stable corner from the contour, then it fits the supporting region of the contour near the corner to an angle, and uses its bisector as the direction of the feature. Next, it searches the contour for the tangent point in the direction of the angle bisector. Finally, with the corner as the centre, and in combination with the tangent point and the feature direction, an elliptic invariant region is constructed. The feasibility of the algorithm was verified experimentally by comparing its repetition rate. Test images obtained from actual scenes include several types of transformations, such as rotation, scaling, affinity, illumination and noise. The results of the experiment show the feasibility of the proposed method for use in local invariant features detection.

Automatic assessment of retinal vessels plays an important role in the diagnosis of various eye, as well as systemic diseases. A public screening is highly desirable for prompt and effective treatment, since such diseases need to be diagnosed at an early stage. Automated and accurate segmentation of the retinal blood vessel tree is one of the challenging tasks in the computer-aided analysis of fundus images today. We improve the concept of matched filtering, and propose a novel and accurate method for segmenting retinal vessels. Our goal is to be able to segment blood vessels with varying vessel diameters in high-resolution colour fundus images. All recent authors compare their vessel segmentation results to each other using only low-resolution retinal image databases. Consequently, we provide a new publicly available high-resolution fundus image database of healthy and pathological retinas. Our performance evaluation shows that the proposed blood vessel segmentation approach is at least comparable with recent state-of-the-art methods. It outperforms most of them with an accuracy of 95% evaluated on the new database.

According to biological attention mechanism, a region of interest (ROI) detection based on visual attention model is closer to human visual system. Taken into account the characteristics of pornographic image during regions detection, a pornographic image region detection method based on visual attention model in compressed domain is proposed in this study, which includes the following four steps: (i) the skin colour regions of pornographic images are detected in compressed domain; (ii) visual saliency map in compressed domain is computed to construct visual attention model; (iii) threshold segmentation method is used for visual saliency map, and then the torso information is retained as pornographic regions; and (iv) four features of colour, texture, intensity and skin are extracted to represent pornographic region. The experimental results show that the proposed method can perform well on the speed/accuracy of pornographic regions detection and representation.

There is an everyday increase of retinal images in different application such as human recognition and diagnosing the eye diseases. Therefore, the need for an automatic method which can recognise the eye parts of retinal images as eye features is unavoidable. This paper offers an automatic and quick morphological-based blood vessel extraction algorithm from the coloured retinal images using Curvelet transform (CT) and principle component analysis (PCA) is proposed. In this algorithm, by estimating the illumination of background and the distribution of contrast in the retinal images, the brightness of images is considerably uniformed. Furthermore, CT is used to enhance the contrast of retinal images by highlighting the edge images in various, scales and directions. We use an improved morphology function introduced with multi-directional structure elements, to extract the blood vessels from retinal images. Connected component analysis and an adaptive filter are used to refine appeared frills with the size of smaller than arterioles in images. The proposed algorithm is evaluated on available images of the DRIVE database and accuracy rate of 94.58% for the blood vessel extraction is obtained. The obtained results show efficiency of the proposed algorithm in comparison with the presented approaches in the literature.

Human vision is highly sensitive to perceive the contrast for interpreting images, so that the neurons of the brain will be excited to make recognition of the surroundings. In this study, the test patterns were inspired to reach the neurons excitation via the proposed pixel resonance concept, and then the quantum concept of physics was employed with random walk behaviour to analyse the test patterns for multiple interpretations. In simulation, similar alphabetic letters (C, O) and Chinese characters (玉, 王) are selected. Subsequently, various extents of noise are added including the different mean values and related standard deviations for the two test patterns. The results showed that the corresponding resonance maps with various morphologies can generate multiple resonance outputs. Furthermore, the simulation results indicated that the peak signal-to-noise ratio curves match those of the resonance maps perceived by using human vision. Based on this approach, it is concluded that the proposed pixel resonance algorithm can effectively simulate the multiple interpretations of the human vision system. Further, a novel physical model was presented to generate multi-output patterns for a single image.