Bekir Dizdaroglu

An improved motion-compensated restoration method for damaged color motion picture films

In this paper, we propose an improved motion-compensated restoration method for color motion picture films deteriorated due to flashing blotches. The method consists of an improved multiresolution block matching with log-D search, a rank ordered differences-based blotch detection and 3D vector median filtering for interpolation of missing data, and utilizes five consecutive frames. Performance of the method is tested on artificially corrupted image sequences and real motion picture films, and is compared to that of the three-frame-based method which involves similar algorithms except improved motion estimation and blotch detection. The results show that the method efficiently works even in severely blotched and motion regions in image sequences.

An improved method for color image editing

In this article, an improved approach for editing color images has been presented. In this approach, using the Poisson equation, a guided vector field is created by employing source and target images within a selected region at the first step. Next, the guided vector is used in generating the result image. Most of the existing techniques in the literature perform image editing without taking color information into account. However, without utilizing color information, image geometry cannot be created properly in some cases which may result in unsatisfactory results. Unlike the existing methods, the proposed study utilizes all the information contained in each of the color channels in computing gradient norm and performing inpainting process. The test results show that the suggested technique generates satisfactory results in editing color images.

A spatiotemporal algorithm for detection and restoration of defects in old color films

A spatiotemporal method is presented for detection and concealment of local defects such as blotches in old color films. Initially, non-local means (NL-means) method which does not require motion estimation is used for noise removal in image sequences. Later, the motion vectors that are incorrectly estimated within defect regions are repaired by taking account of the temporal continuity of motion trajectory. The defects in films are detected by spike detection index (SDI) method, which are easily adapted to color image sequences. Finally, the proposed inpainting algorithm fills in detected defect regions, which is not required to estimate true motion like other approaches. The method is presented on synthetic and real image sequences, and efficient concealment results are obtained.

An Image Completion Method Using Decomposition

This study presents a hybrid technique for simultaneously completing images by using geometry and texture components of input data. The approaches using inpainting methods based on partial differential equations (PDEs) to fill in large image regions usually fail if these regions contain textures. On the other hand, texture synthesis algorithms sometimes fail due to complex structures and textures in the image. However, this study, suggesting a hybrid method using both techniques, produces satisfactory results in completing the missing parts of images. In the proposed method, the given image is decomposed into two components. The geometry component, obtained by using the regularization PDE based on a trace operator, was inpainted by a tensor-driven PDE algorithm that takes curvatures of line integral curves into account, and the texture component, obtained by subtracting the given image from the geometry component, was reconstructed by the modified exemplar-based inpainting algorithm. Both of these methods work on color information. The main contribution of this paper is that it uses decomposition and montage stages together which provides superior results compared with the existing methods. Experimental results show that the proposed method efficiently fills in target regions, which is promising.

Variational additive noise removal of color images

In this study, in color images additive noises which appear as corruptive effects are reduced. In the proposed method, original structure and texture of the input image is preserved well while the noise is removed from the image. In other words, in the proposed method, energy functional used in the noise removal process includes both smoothing term and fidelity term. In addition, these terms are optimized at each iteration, depending on the type of noise and its statistical information. The proposed method generates efficient results in terms of both quality and quantity.

Detection of painting cracks using optimal thresholding

In this study, the existing optimal image thresholding techniques are reviewed for detection of painting cracks in general. Cracks on the painting can be detected by using the morphological top-hat transformation. However, since there are quite a number of false detections, a thresholding method should be applied to the obtained result. Therefore, optimal thresholding methods such as Otsu, Minimum error and Fuzzy thresholding can be used for overcoming this problem. Experimental results demonstrate the validity of the thresholding methods.

Automated removal of Gaussian noise in color images

In this study, an automatic trace-based approach is presented for reducing Gaussian noise in color images. As a first step in the method, homogeneous regions in the noisy image are detected. As a second step, the variance of additive white Gaussian noise is estimated by taking into account the detected homogeneous regions. As a final step, the noise reduction process is performed via an improved trace-based approach that considers both structure and texture information in the color image. The obtained results depict the performance of the method.

Medical image segmentation with edge based level sets: Mobile client-server application

In this study, segmentation of medical images is implemented by an edge-based level set approach based on the client-server communication. In the method, a medical image, which is already loaded into the image gallery of the smartphone by the user, is selected, the pathological region to be segmented is roughly marked and then both the input image and the marked region, and some input parameters used in the segmentation process are sent to the local server. After the segmentation process is performed in the local server, the resulting image is downloaded again to the mobile device. Since the segmentation process is highly dependent on the edge information that must be robustly obtained, the Canny edge detection method, which includes some improvement, instead of classical forward finite differences is used for the edge indicator and thus the performance of the method is increased. Obtained visual results also support this situation.

Image inpainting based on structure and texture components

In this study, the input image is decomposed into structure and texture components and the trace based and patch based inpainting approaches are applied to the obtained components respectively. Then the generated results are added together and the resulting image is obtained. The proposed method is compared with some of the inpainting methods in the literature. Structure and texture information in the region which is inpainted are estimated by the proposed method in a better way when compared with other methods.

Client-server based photomontage

This study presents a client-server architecture adopted for solution of the Poisson equation, which is a linear equation, used in photomontage. Firstly, a general idea is given about main principle of working of the proposed approach. Secondly, an approach taking into account iterative methods in order to solve the Poisson equation based on Dirichlet boundary conditions is presented. After that it is given a general idea about remote method invocation (RMI) which is used to achieve a client-server architecture in Java programming language and how the RMI approach is put into practice for photomontage. Lastly, tests and generated results are discussed.