A New Efficient Image Inpainting Method

This study proposes a novel image inpainting technique based on watermarking and halftoning. The LSB method is used to embed error diffusion halftone image into original image for protecting the image. In image repair process, we use LSB method to extract the halftone information, and the reference image is achieved from LUT inverse halftone. Finally, the reference image is used to finish the image inpainting work. Experimental results show the performance of our method is very excellent in image inpainting

The thesis is based upon Taiwanese temple painter, Li, Deng Sheng, as the central subject, which analyzes various periods of temple painting developments, as well as the heritage conditions of temple painters from the north, central and south regions of Taiwan; meanwhile, categories and subject matters of his works will be provided with detailed descriptions, including different stages of apprenticeship of Li, such as learning the craft in temples to making a living on some skill, difficulties and dilemma he suffers and learns from; as well as his self-promoting and transitions. On the other hand, through issues of social networks and identity recognition, the balance Li attempts to obtain in practical economy, ideals and dreams, as well interactions with others will be analyzed, too. Still further, from his temple painting works, the thesis contains an attempt to provide understanding of how work pieces from current cultural crafts and their preservers would present differently from earlier times using digital collecting technologies. Nonetheless, due to differences of viewers and quests, varieties often generates between images in temple painting and classic documents. In that case, an example of portraits of Su Shi painted by Li shall be offered to both inspect and compare with historical data, as well as further analyze how Li attempted to figure the Su’s enthusiasm as a literatus, how Su interacted with Fo Yin, and a trip in the Red Cliff. Eventually, issues like operating painted temples and heritage shall be offered, and provided with suggestion and direction for researches; it is in the hope of benefitting relevant researchers that tend to study same genre, and of expecting to deliver the concepts of cooperation between generations to accomplish the responsibilities of the temple painting heritage through offering crafting skills and experience of the former generation.

Tarquinian tomb paintings are of an unrivalled artistic and cultural importance. The primary aim of this thesis is to identify and investigate their pictorial forms, language, and structures of painted dress, so as to re-tailor understanding of Tarquinia, its tombs, and its tomb paintings throughout the Archaic and Classical periods of ancient Italy (approx. 6th – 2nd century BC). Tarquinian painted dress is examined by the application of an original methodological approach, which identifies and investigates the structures of painted dress through the application of a hierarchical clustering analysis. Thereby, the thesis emphasises the significant versatility of dress as an investigative tool in the ancient world, and expands its study, and extends its forms of analysis. Methodologically, it innovatively quantifies Tarquinian tomb painting, which contrasts the typical qualitative and anecdotal methods usually applied when dealing with such visual imagery depicted amidst archaeological contexts. The analysis makes use of numerical procedures to divide a group of given units into homogeneous sub-groups, which allow for the visual inspection of an otherwise complex set of data-matrices. This statistically examines painted dress’s relationships to identify groupings of similarity from within a heterogeneous data-matrix. Seeking deeper structures not readily apparent on the surface, or revealed by more traditional modes of analysis, the study facilitates identification of painted dress’s arrangements, and examines their groupings by applying key tenets of contemporary dress theory. Trends of typological regularity, high frequency, but weak associative strength belonging to 6th – 5th century BC painted dress groupings are identified, as are trends of typological irregularity, low frequency, but strong associative strength, identified as belonging to 4th – 3rd/2nd century BC groupings. Thus, this study contributes a hitherto unidentified varying manipulation of non-verbal sartorial communication, which indicates consciously shifting choices, priorities, intentions, and decisions as to the communicative utilization of painted dress from the 6th – 3rd/2nd century BC. Therefore, painted Tarquinian dress is identified as a key form of non-verbal communication that conveys previously overlooked socio-cultural information, but the thesis also repositions tomb painting as a communicative device. Consequently, it provides new insight into Etruscan dress’s and painted tombs’ broader significance amidst wider Etruscan society.

Although much has been written about the art of the famous synagogue at Dura-Europos, its rootedness in Mesopotamia has gone largely unexplored. This study looks south along the local trade routes to Iranian Babylonia and examines evidence available about the religious function of Durian Jewish and Sasanian Manichaean pictorial art as part of a shared regional development of techniques of instruction. It reveals that the distinctly different forms of pictorial art used by these two communities in mid-third-century Mesopotamia are nevertheless comparable based on their didactic function. They both: (1) displayed a visual library of doctrinal subjects, that is, they captured, in pictorial form, a large sample of core tenets which were also recorded in the respective sacred texts of these religions; (2) fulfilled a primarily didactic function, that is, their pictorial genres (narrative scenes, didactic portraits, and diagrams in the Manichaean case) played a dominantly instructional role; and (3) effectively supplemented oral instruction, that is, the paintings were sermonized about and discussed in light of living interpretations. I argue that these correlations result not from direct influence between the two communities, but rather from a shared approach to what images can do for a religion. The Jewish and Manichaean paintings in question emerged simultaneously and in relative closeness to one another. While the Jewish archeological records of the painted synagogue are all but silent, various characteristics of the mid-third-century Manichaean paintings are noted in literary records, including what they portrayed and, most importantly for this study, the pedagogical reasons for how and why they were used. As evidenced by Iranian, Coptic, and Syriac textual sources from between the mid-third and the late fourth and early fifth centuries, the founding prophet of Manichaeism, Mani (active from 240 to 274/277 CE), not only wrote down his own teachings, but also created visual representations of them on a solely pictorial scroll—the Book of Pictures—that he and his highest-ranking elects used in the course of oral instructions while missionizing across greater West Asia and the East Mediterranean region.

In the field of computer assisted diagnostic, Image segmentation is one of the most important preprocessing steps. In this paper, the watershed algorithm segmented image by choosing different merging parameters, original image will be divided into different scales of image, High-scale image and low-scale image. High-scale image is part of the pulmonary parenchyma without the main vessel in it, low-scale image used to aid region growing tore move the branching vessels from the pulmonary parenchyma. Selected seed point located inside the branch vessels and determined the region growing threshold, with the aid of low-scale image region growing can segment the branch vessel regions .Removed the regions in the high-scale image to get the final segmentation results. In this paper, a large number of CT lung data segmentation experiment. Proven, good segmentation results are achieved.

Millions of digital images are captured by imaging devices on a daily basis. The way imaging devices operate follows an integral process from which the information of the original scene needs to be estimated. The estimation is done by inverting the integral process of the imaging device with the use of optimization techniques. This linear inverse problem, the inversion of the integral acquisition process, is at the heart of several image processing applications such as denoising, deblurring, inpainting, and super-resolution. We describe in detail the use of linear inverse problems in these applications. We review and compare several state-of-the-art optimization algorithms that invert this integral process. Linear inverse problems are usually very difficult to solve. Therefore, additional prior assumptions need to be introduced to successfully estimate the output signal. Several priors have been suggested in the research literature, with the Total Variation (TV) being one of the most prominent. In this thesis, we review another prior, the l0 pseudo-norm over the gradient domain. This prior allows full control over how many non-zero gradients are retained to approximate prominent structures of the image. We show the superiority of the l0 gradient prior over the TV prior in recovering genuinely piece-wise constant signals. The l0 gradient prior has shown to produce state-of-the-art results in edge-preserving image smoothing. Moreover, this general prior can be applied to several other applications, such as edge extraction, clip-art JPEG artifact removal, non-photorealistic image rendering, detail magnification, and tone mapping. We review and evaluate several state-of-the-art algorithms that solve the optimization problem based on the l0 gradient prior. Subsequently we apply the l0 gradient prior to two applications where we show superior results as compared to the current state-of-the-art. The first application is that of single-image reflection removal. Existing solutions to this problem have shown limited success because of the highly ill-posed nature of the problem. We show that the standard l0 gradient prior with a modified data-fidelity term based on the Laplacian operator is able to sufficiently remove unwanted reflections from images in many realistic scenarios. We conduct extensive experiments and show that our method outperforms the state-of-the-art. In the second application of haze removal from visible-NIR image pairs we propose a novel optimization framework, where the prior term penalizes the number of non-zero gradients of the difference between the output and the NIR image. Due to the longer wavelengths of NIR, an image taken in the NIR spectrum suffers significantly less from haze artifacts. Using this prior term, we are able to transfer details from the haze-free NIR image to the final result. We show that our formulation provides state-of-the-art results compared to haze removal methods that use a single image and also to those that are based on visible-NIR image pairs.

A simple and efficient watermark method is proposed in this paper. The watermark method is an excellent technique to protect the copyright ownership of a digital image. The proposed watermark method is built up on the concept of visual cryptography. According to the proposed method, the watermark pattern does not have to be embedded into the original image directly, which makes it harder to detect or recover from the marked image in an illegal way. It can be retrieved from the marked image without making comparison with the original image. The notary also can off-line adjudge the ownership of the suspect image by this method. The watermark pattern can be any significant black/white image that can be used to typify the owner. Experimental results show that the watermark pattern in the marked image has good transparency and robustness. By the proposed method, all the pixels of the marked image are equal to the original image.

Breast cancer continues to be one of the main causes of death among women. Various studies have confirmed that the early detection of sub-clinical cancers may improve the prognosis. X-ray mammography in this case is the best diagnostic technique. It’s based on the interaction of a cone beam X-ray with the mole tissue. The projection image obtained can be analyzed qualitatively by the radiologists. But, an automatic treatment and quantitative analysis of this kind of images is suitable. For this reason several studies are conducted to develop tools to help with diagnosis of this disease (CAD: Computer-Assisted Diagnosis). We propose in this paper a new method to segment mammographic images based partly on a pyramidal architecture. The original image is fragmented (quadtree) initially to homogeneous regions. Each region is then associated with a peak of graph. It gathers data within homogeneous groups named regions classes’ c, then we use HCA (Hierarchical classification ascendant) and k-means to find the optimal partition for the largest possible value of c at the initial stage. This technique gives good results, and allows calculating morphological parameters of the breast cancer.

Cucumber fruit appearance quality is an important basis of growth status. In order to improve the quality detection accuracy and processing efficiency of cucumber color image under complicated background, an improved GrabCut algorithm was proposed to extract the cucumber boundary. Firstly, including pixel size normalization, rectangular box set and scale image resolution, pretreatments of cucumber image were adopted to reduce the iteration times and operation time of GrabCut algorithm. Then, the Gaussian mixture model was chosen to find out the possible prospect of target region and background region in the preprocessed rectangular frame on the preliminary modeling. Meanwhile, by the optimization of K-means cluster to the initial GMM model, the effective target area was extracted. Finally, the whole image noise and serrated boundary was removed by morphological operations to segment the outline of the complete target prospects with appropriate structure size. And then the cucumber appearance quality detection instrument was designed to extract the texture and shape features exactly, so that it could obtain cucumber appearance quality and evaluate its growth effectively. With the segmentation experiments by almost 300 cucumber original images from greenhouse in Shandong Province, the results showed that the improved GrabCut algorithm could effectively extract the complete and smooth boundary of cucumber. With relatively high segmentation evaluation index, the precision was 93.88%, the recall rate was 99.35%, the F-Measure reached 96.53%, and the misclassification error was controlled at minimum 5.84%. The average running time was shortened to 1.4023 s. The comparison results showed that the improved GrabCut algorithm was the best, followed by the SLIC and traditional GrabCut method. Cucumber appearance quality detection instrument could also extract more accurate feature parameters. And it could meet the basic growth condition assessment by automatic image processing. Keywords: cucumber, complicated background, quality detection, image processing, GrabCut DOI: 10.25165/j.ijabe.20181104.3090 Citation: Ye H J, Liu C Q, Niu P Y. Cucumber appearance quality detection under complex background based on image processing. Int J Agric & Biol Eng, 2018; 11(4): 193-199.

For health and safety monitoring in civil constructions such as bridges, roads and pavements, segmenting the regions of interest is the fundamental requirement for image analysis at high-level semantic. One of the major structural problems in concrete and asphalt surfaces are cracks, which start with harming the visual aspect of the construction and further lead to failure of the construction. This study makes four contributions to extract the image foreground from the background in order to address the crack detection task on the asphalt and concrete surfaces. Specifically, we model cracks as foreground objects and concrete or asphalt surfaces as textured background. The first contribution of this research is a hybrid image processing method for crack detection. In this method, the cracks are modelled as linear structures on the background of textured concrete or asphalt surfaces, which can be extracted by combining structure extraction with global pattern distribution. There are two phases in this model towards creating the final structure-texture map. The first phase is extracting strong structures or edges using relative total variation measures, which produces a structure feature map by preserving the edges and suppressing the background noises. The second phase calculates the spatial distribution of textures across the image. A bag-of-words model is used in this phase to quantise the texture pattern, which in crack detection application is the widely distributed road texture background. The local structure map and the global distribution map extract the crack structure and the textured background, respectively. The final crack is extracted by fusing these two maps and applying binarisation as the post-processing step. This model achieved a better result compared with the local structure extraction and the saliency method. As the second contribution, a large-scale dataset of asphalt and concrete crack images is prepared, including images with their corresponding high-resolution pixel-wise labels. To the best of our knowledge, until the completion of this thesis, there has not been any crack image dataset with pixel-wise ground truth labels available. The original crack image set includes 2532 images of cracks on brick and asphalt surfaces. This image set is split into training, validation and testing sets with the ratio of 50/25/25. Two augmentation techniques of rotation and ipping are applied to only the training set while validation and testing sets are locked in order to prevent the data leakage. All models are learned on the training set after fine-tuning the hyper-parameters. After each tuning, an early estimate of the model accuracy is obtained using the validation set. Finally, an unbiased performance estimation of the fitted model is provided on the testing set. The third contribution of this thesis is developing two encoder-decoder networks by exploring the recent advances of deep learning research for crack detection and applying them on our crack dataset. The first network is inspired by DeepLab, which is a modified ResNet architecture. In this network the last pooling layer is replaced with an Atrous Spatial Pyramid Pooling (ASPP) module. This encoder-decoder structure is designed to classify each image pixel into two foreground cracks or textured background. The second model is inspired by Full Resolution Residual Network (FRRN), which is a ResNetlike network with two residual and pooling streams to extract the high- and low-level features, respectively. The combination of di erent level features in this model improves the localization of crack pixels as well as the recognition of the crack structure as a whole. As FRRN outperforms DeepLab on crack classification, we select it as the baseline for further research. In our last contribution, we optimise the FRRN model by reducing the number of parameters. Inspired by the Inception module which significantly improved the utilization of the computing resources inside the GoogLeNet, we proposed Incepted FRRN (I-FRRN) network by embedding the Inception module inside the FRRN. Combining these two structures, our proposed model records 88.14% accuracy in classifying the positive class with 0.22% improvement, while having less than half the number of parameters compared with FRRN. The results show that the proposed architecture achieves significant computational e ciency gains and comparable or higher class-accuracy in crack classification task over the baseline model.

This article considers the collections of painters and the invention of the typical Antwerp-genre of the gallery picture. It holds that these pictures were a ‘defense of the image’ and must be understood in Antwerp’s Counter Reformation context, in the aftermath of the iconoclastic furies of the sixteenth century. Quintessential to both iconoclasm and the culture of collecting were issues of display and materiality. The paintings of cabinets suggest that some Antwerp artists were defenders of the image in a time when images and other man-made objects were at the forefront of intellectual debate. The gallery paintings can be related to a widespread debate on materiality that, implicitly or explicitly, extended from scholarly publications to workshops and collectors’ rooms.

Hydrometeological forcing data with high spatio-temperal resolution,such as wind,temperature,pressure,humidity,radiation,precipitation,is needed for driving regional-scale and watershed-scale hydrological,ecological,land models or data simulations system.In this paper,the importance of regional-scal and watershed-scale forcing data is clarified;methods of producing forcing data,such as interpolation based on observation data,reanalysis data,dynamical downscaling method,statistical downscaling method and multi-source information fusion method,are described in detail;some cases about precipitation,radiation,temperature and other variables are given;some case studies on producing forcing data for land data assimilation systems are introduced,such as Global Land Data Assimilation System(GLDAS),European Land Data Assimilation System to predict Floods and Droughts(ELDAS),North American Land Data Assimilation System(NLDAS),Chinese Land Data Assimilation System(CLDAS) and Land Information System(LIS).Finally,existing problems of producing high spatio-temperal forcing data for regional-scale and watershed-scale hydrological,ecological,land models or data simulations system are pointed out in detail,and four proposals are put forward: ①it is difficult to satisfy all forcing variables by one simple method;②it is urgent to provide proper land surface initial condition for dynamical downscaling model;③it is a major method to combine with Doppler radar data,drop size meter and rain gauge observations to improve the accuracy of rainfall;④it is a trend to input remote sensing images into dynamical downscaling model by data assimilation method to generate high spatio-temperal forcing data.Some ideas on preparing forcing data for land surface model of Western China are proposed.

The way we experience space has a direct relationship to the way we perceive it, as evidenced by the ways that space has been represented in painting throughout history. My research is concerned with the representation of space in contemporary painting. Contemporary experiences of space through new media screens offer painters a unique challenge that requires them to think about representing space in new ways. My research focuses on the role of the window in painting, a device that has confirmed painters’ preoccupation with representing space on a two-dimensional plane. I provide an historical overview that establishes the window as an important spatial and metaphorical concern within painting. I draw a connection between the window and Plato’s cave as a frame of representation. In the context of this research, ‘space’ refers to a painterly space which includes both illusory space and actual physical/material space. Whereas Gilles Deleuze defines these different kinds of space as either ‘haptic’ or ‘optic space’, I use the term ‘unstable space’ to describe that which occurs when both ‘haptic’ and ‘optic space’ coexist on a picture plane. In considering Plato’s cave as a window or frame of representation, I recognise the demarcation of key spatial and representational concepts related to the window in painting. As with Plato’s cave, the window demarcates binary opposites that have structured much subsequent thinking about art, such as interiority/exteriority, nature/culture, illusion/reality. Through Plato’s theories, I specifically draw attention to the dualist structure of his belief system that posits tensions between interior and exterior, reality and illusion, nature and culture. I establish the term ‘unstable space’ through examining the theories of Deleuze and Jacques Derrida that deconstruct Plato’s writing. Their theories offer a means by which to construct new meanings within the space of painting, as they emphasise the instability of the binaries afforded by Plato’s philosophy and instead suggest the possibilities of multiplicity. The contemporary ‘windows’ of media screens significantly shift the metaphor of the singular window to the multiplicity of windows within windows. This multiplicity reflects the way we currently experience space and the effect this has on the thinking of space in contemporary painting practice. In this way, painting’s material dimension and illusory space can be explored not in terms of binary oppositions but as complementaries. Having traced the development of the window as an important representational device in painting, I propose that the window can be used as a mechanism to explore the ‘unstable space’ through painting. This space operates between spatial and representational theory. Through the analysis of specific works of art by contemporary artists chosen as exemplars in the field of painting, I argue that the unstable space can be created solely in the medium of paint. My research extends understandings of space as represented within the limits of a two-dimensional surface. The representation of space within my painting practice results from my reimagining of the window as an unstable space, my exploring of the perception and representation of ambiguous space, and my engaging with pictorial illusion through abstraction. Explicitly, the studio research found that the screen or window was able to act as a metaphor for the body and as such effectively articulate the experience of interiority and exteriority, surface and figure, ground and distance.

information from the original image as compared with crisp or hard segmentation methods. In practice, noisy images (even high noise images) are very co mmon. It's very essential and critical to deal with such images to process real-image segmentation and pattern recognition. In this paper, differences of credibilistic clustering algorithm (CCA) and fuzzy c-means algorithm (FCM) in dealing with noisy images are studied and the research shows that in most cases, CCA performs better than FCM in high noise image segmentation. Based on that, a new kind of fuzzy clustering methods is presented. It combines spatial credibilistic clustering algorithm (SCCA) with particle swarm optimization (PSO) and takes full advantages of them. The advantages that come from CCA in noise image segmentation also help in SCCA, and the imposition of spatial information enlarges the advantage. The addition of PSO helps to improve global search performance; thereby the novel methods overcome the drawback of single clustering methods - local optimal solutions. Computational experiments show that the proposed methods give the best segmentation results when compared with FCM, CCA, spatial fuzzy c-means algorithm (SFCM), SCCA and the PSO incorporated versions of FCM, CCA, and SFCM.

A combination of dual tree complex transform and the image processing algorithm of semi-blind deconvolution was proposed to eliminate the factors which make the image worse. Firstly, the retinal image obtained from the adaptive optics system was decomposed with dual tree complex transform into two parts. The low frequency image was processed by the algorithm of semi-blind deconvolution with some constraints. And the optical transfer function was used as initial parameter estimate which was constructed with the residual aberration of image system. The high frequency image was processed by denoising. Finally, the goal image was obtained by combining the processed images. The experimental results show that the retinal image quality is improved with this method, the image objective quality evaluation parameters are increased more than 5 times compared with the original image, and the average power spectrum is improved about 6 times in the spatial frequency range(70~90()-1) of retinal cells, which show that the method can contribute to satisfying the requirement of observation of human retinal image.