Sharif Digital Repository

In this paper, we propose an adaptive object tracking algorithm in crowded scenes. The amplitudes of of Undecimated Wavelet Packet Tree coefficients for some selected pixels at the object border are used to create a Feature Vector (FV) corresponding to that pixel. The algorithm uses these FVs to track the pixels of small square blocks located at the vicinity of the object boundary. The search window is adapted through the use of texture information of the scene by finding the direction and speed of the object motion. Experimental results show a good object tracking performance in crowds that include object translation, rotation, scaling and partial occlusion. Copyright - World Automation... 

Few-view or sparse-view computed tomography has been recently introduced as a great potential to speed up data acquisition and alleviate the amount of patient radiation dose. This study aims to present a method for high-quality ultra-fast image reconstruction from sparsely sampled projections to overcome problems of previous methods, missing and blurring tissue boundaries, low-contrast objects, variations in shape and texture between the images of different individuals, and their outcomes. To this end, a new deep learning (DL) framework based on convolution neural network (CNN) models is proposed to solve the problem of CT reconstruction under sparsely sampled data, named the multi-receptive... 

Real-world image classification, which aims to determine the semantic class of un-labeled images, is a challenging task. In this paper, we focus on two challenges of image classification and propose a method to address both of them simultaneously. The first challenge is that representing images by heterogeneous features, such as color, shape and texture, helps to provide better classification accuracy. The second challenge comes from dissimilarities in the visual appearance of images from the same class (intra class variance) and similarities between images from different classes (inter class relationship). In addition to these two challenges, we should note that the feature space of... 

Processing time and segmentation quality are two main factors in evaluation of image segmentation methods. Oversegmentation is one of the most challenging problems that significantly degrade the segmentation quality. This paper presents a framework for decreasing the oversegmentation rate and improving the processing time. Significant variations in both color and texture spaces are the main reasons that lead to oversegmentation. We exploit Affine Invariant Region Detectors to mark regions with high variations in both color and texture spaces. The results are then utilized to reduce the oversegmentation rate of image segmentation algorithms. The performance of the proposed framework is... 

Recently, a technique has been proposed for image hiding, that is based on block texture similarity where, blocks of secret image are compared with blocks of a set of cover images and the cover image with the most similar blocks to those of the secret image is selected as the best candidate cover image to conceal the secret image. In this paper, we propose a new image hiding method in which, the secret image is initially coded using a fuzzy coding/decoding method. By applying the fuzzy coder, each block of the secret image is compressed to a smaller block. In this way, after compressing the secret image to a smaller one, we hide it in a cover image. Obviously hiding a smaller secret image... 

We present a modified examplar-based inpainting method in the framework of patch sparsity. In the examplar-based algorithms, the unknown blocks of target region are inpainted by the most similar blocks extracted from the source region, with the available information. Defining a priority term to decide the filling order of missing pixels ensures the connectivity of object boundaries. In the exemplar-based patch sparsity approaches, a sparse representation of missing pixels was considered to define a new priority term. Here, we modify this representation of the priority term and take measures to compute the similarities between fill-front and candidate patches. Comparative reconstructed test... 

Watermarking is a suitable approach for digital image authentication. Robustness regarding attacks that aim to remove the watermark is one of the most important challenges in watermarking, in general. Several different attacks are reported that aim to make it difficult or impossible for the real owner of the digital watermarked image to extract the watermark. Some of such common attacks are noise addition, compression, scaling, rotation, clipping, cropping, etc. In this paper we address the issue of print and scan attack by introducing a method to model the scanner and printer. Then we will simulate the print and scan attack on the digital images to evaluate its robustness. In addition, we... 

Partial discharge (PD) measurement is one of the best methods for condition monitoring of transformers. In this paper, we use 5 different types of defects as follows: scratch on winding insulation, bubble in oil, moisture in insulation paper, a very small free metal particle in the transformer tank, and a fixed sharp metal point on the transformer tank, for our PD-related studies. Each type of defect is implemented into 1 of the 5 identical transformer models, which had been developed in the authors' recent work. The continuous wavelet transform is applied to each related measured time-domain PD signals. This process results in an image, for each PD pulse in the time-frequency domain. Using... 

Suggestion and application of a set of new features for on-line Partial Discharge (PD) monitoring, where there is no information about the type of PD is a challenging task for condition assessment of power equipments, such as a power transformer. This is looked for in this paper. So far, in past various techniques have been employed to develop a comprehensive PD monitoring system, however limited success has been achieved. One of the challenging issues in this field is the discovering of proper features capable of differentiating the involvement of possible types of PD sources. In order to examine the efficiency of the method established in this paper, which is based on application of a set... 

Skin detection is a difficult and primary task in many image processing applications. Because of the diversity of various image processing tasks, there exists no optimum method that can perform properly for all applications. In this paper, we have proposed a novel skin detection algorithm that combines color and texture information of skin with cellular learning automata to detect skin-like regions in color images. Skin color regions are first detected, by using a committee structure, from among several explicit boundary skin models. Detected skin-color regions are then fed to a texture analyzer which extracts texture features via their color statistical properties and maps them to a skin... 

Feature vector extraction, based on local image texture, is a primitive algorithm for many other applications, like segmentation, clustering and identification. If these feature vectors are a good match to the human visual system (HVS), we can expect to get the appropriate results by using them. Gabor filters has been used for this purpose successfully. In this paper we introduce a novel refinement, with the use of Matching Pursuit (MP) to improve the Gabor based texture feature extractor. With this improvement, we show that the separability of different textures will increase. Another consideration in this work is computation complexity. Therefore, we limit the basis function set to reduce... 

Content-based image retrieval (CBIR) has been one of the most active areas of research. The retrieval principle of CBIR systems is based on visual features such as colour, texture and shape or the semantic meaning of the images. A CBIR system can be used to locate medical images in large databases. This paper presents a CBIR system for retrieving digital human brain magnetic resonance images (MRI) based on textural features and the support vector machine (SVM) learning method. This system can retrieve similar images from the database in two groups: normal and tumoural. This research uses the knowledge of the CBIR approach to the application of medical decision support and discrimination... 

Content-based image retrieval (CBIR) has turned into an important and active potential research field with the advance of multimedia and imaging technology. It makes use of image features, such as color, texture and shape, to index images with minimal human intervention. A CBIR system can be used to locate medical images in large databases. In this paper we propose a CBIR system which describes the methodology for retrieving digital human brain magnetic resonance images (MRI) based on textural features and the Adaptive neuro-fuzzy inference system (ANFIS) learning to retrieve similar images from database in two categories: normal and tumoral. A fuzzy classifier has been used, because of the... 

The quasi-static and strain-controlled fatigue characteristics of ZK60 extrusion have been investigated along three different directions: the extrusion direction (ED), the radial direction (RD), and 45° to the extrusion direction (45°). The quasi-static response showed symmetric behavior for the samples tested along RD and 45° whereas the ED samples manifested completely asymmetric behavior. Although the ED samples exhibited longer fatigue lives than the RD and 45° in the high cycle fatigue, the fatigue lives in the low cycle fatigue regime were similar. The texture measurement indicated a sharp basal texture along ED, explaining its asymmetric behavior. Higher tensile mean stress and less... 

Retained austenite as a key constituent in final microstructure plays an important role in TRansformation Induced Plasticity (TRIP) steels. The volume fraction, carbon concentration, size, and morphology of this phase are the well-known parameters which effects on the rate of transformation of retained austenite to martensite and the properties of steel, are studied by many researchers. Of the transformation of retained austenite to martensite under strain in a TRIP steel is studied in this paper. The experimental results show that the transformation rate of retained, austenite with similar characteristics, to martensite in differently processed TRIP steel samples, exhibits an anisotropic... 

The problem of motion detection has received considerable attention due to the explosive growth of its applications in video analysis and surveillance systems. While the previous approaches can produce good results, the accurate detection of motion remains a challenging task due to the difficulties raised by illumination variations, occlusion, camouflage, sudden motions appearing in burst, dynamic texture, and environmental changes such as weather conditions, sunlight changes during a day, and so on. In this paper, a novel per-pixel motion descriptor is proposed for motion detection in video sequences which outperforms the current methods in the literature particularly in severe scenarios.... 

We present an attempt to improve the performance of multi-class image segmentation systems based on a multilevel description of segments. The multi-class image segmentation system used in this paper marks the segments in an image, describes the segments via multilevel feature vectors and passes the vectors to a multi-class object classifier. The focus of this paper is on the segment description section. We first propose a robust, scale-invariant texture feature set, named directional differences (DDs). This feature is designed by investigating the flaws of conventional texture features. The advantages of DDs are justified both analytically and experimentally. We have conducted several... 

The main concern of steganography (image hiding) methods is to embed a secret image into a host image in such a way that the host should remain as similar as possible to its original version. In addition the host image should remain robust with respect to usual attacks. In this paper we present a method that tries to cover all above mentioned concerns. The secret and host images are divided into blocks of size 4 × 4. Each block in secret image is taken as a texture pattern for which the most similar block is found among the blocks of the host image. The embedding procedure is carried on by replacing these small blocks of the secret image with blocks in host image in such a way that least... 

Segmentation of aerial images has been a challenging task in recent years. This paper introduces a simple and efficient method for segmentation and classification of aerial images based on a pixel-level classification. To this end, we use the Gabor texture features in HSV color space as our best experienced features for aerial images segmentation and classification. We test different classifiers including KNN, SVM and a classifier based on sparse representation. Comparison of our proposed method with a sample of segmentation pre-process based classification methods shows that our pixel-wise approach results in higher accuracy results with lower computation time  

Herein, we report a new strategy for improving the efficiency and reducing the fabrication cost of dye-sensitized solar cells (DSCs) by elimination of the three- or four-fold layer deposition of TiO2. This is performed by replacing a single layer deposition of mesoporous TiO2 beads, with sub-micrometer size, high surface area and tunable pore size, synthesized by a combination of sol-gel and solvothermal methods. Furthermore, superior electronic properties gained by a reduction in electronic trap states are achieved through doping of pristine TiO2 beads with lithium. The beads have a spherical shape with monodispersed texture consisting of anatase-TiO2 nanocrystals and ultra-fine pores. The...