Sharif Digital Repository

Despite recent advancements in deep learning methods, these methods rely on a huge amount of training data to work. Recently the problem of solving classification and recently semantic segmentation problems with a few training data have gained attention to tackle this issue. In this research, we propose a meta-learning method by combining optimization-based and prototypical approaches in which a small portion of parameters are optimized with task-specific initialization. In addition to this and designing other parts of the method, we propose a new approach to use query data as an unlabeled sample to enhance task-specific learning. Alongside the mentioned method, we propose an approach to use... 

SAR image noise is a significant problem for SAR image analysis.The inherent noise of SAR images, known as speckle, seriously affects the SAR image interpretation. It also has adverse effects on the classification and segmentation of SAR images. Due to its great significance, the SAR image processing has received considerable attention in recent years and many researchers have developed techniques to reduce the inherent noise accompanying the SAR images. A survey of the literature shows that the wavelet analysis is one of the most common methods used for speckle reduction. While the power of the morphological analysis method has mostly not been recognized, we have utilized this efficient... 

Underground reservoirs are highly complicated due to the presence of spatial heterogeneities at length scales that span from micrometer in pore structure of the rocks to kilometer in the reservoir models. While large-scale flow units need to be characterized using seismic and well data, detailed displacements of fluids in pore space need to be modeled using thin section analysis and pore network modeling. It is therefore necessary to adopt a multi-scale approach to reservoir description to make best use of all the available data that vary over several orders of magnitude, from micro-scale in pore structure to field scale in reservoir flow models. In this thesis, an integrated framework for... 

The aim of this thesis is to investigate recent methods for Image Processing(Any signal process which it’s input is an image and it’s ouput is an image or a set of Image parameters) using Calculus of variation tools. Methods which are to be investigated has been divided into two well known parts of Image Processing : Image Restoration and Image Segmentation.Image Processing Chapter includes two sections: one calculus of variations methods(energy method), other methods based on PDEs(heat equation and Malik-Perona equation). In studing each of this methods, It has been tried to include experimental results and negative and positive points of them.In Image Segmentation Chapter, first... 

Image segmentation is middle and an important task in image analysis and machine vision applications. The output images of imaging systems are often fuzzy because of noise, limitation in spatial and temporal resolution, blurring and intensity inhomogeneity in the objects. The goal of this thesis is exploring the fuzzy methods in multispectral image segmentation and proposing a new one to solve some of the recent difficulties and problems. The difficulties and problems such as simultaneous utilization of spatial and spectral information, necessity for dimension reduction, spatial and spectral and intra-cluster image information redundancy, existence of regions with widely varying size,... 

Image segmentation in medical imaging, as one of the most important branches of medical image analysis, often faces the challenge of limited labeled data for application in deep learning methods. The high cost of data collection and the need for expertise in image segmentation, particularly in three-dimensional images such as MRI and CT or sequence images like CMR, have all contributed to this problem, even for popular networks like U-Net, which struggle to achieve high accuracy. As a result, research efforts have focused on semi-supervised learning approaches, weakly supervised learning, as well as multi-instance learning in medical image segmentation. Unfortunately, each of these methods... 

Environment Identification and recognizing surrounding objects is an exigent need in future applications. For example one of the emerging technologies in car industry is driverless cars. In driverless cars, navigation system should be able to detect and recognize pedestrians, traffic signs, roads, surrounding cars and so on. Therefore, conventional single-object recognition systems are not capable of handling the needs of advanced machine vision based applications. In recent years, designing and analyzing multi-class object detection and recognition systems have become a big challenge in machine vision. In this thesis our goal is to identify and analyze the existing problems in designing... 

For radiotherapy and removal of cancerous tissues, it is necessary to determine the location of the tumor and the vulnerable structures around the tumor before treating and irradiating the high-energy beam. To do this, the images received from the patient need to be segmented. This is usually done manually, which is not only time consuming but also very expensive.Various methods for segmenting these images are presented automatically and semi-automatically, among which methods based on machine learning and deep learning have shown much higher accuracy than other methods. Despite this superiority, these methods have problems such as high computational costs, inability to learn the shape and... 

Due to the overall progress in the processing of imaging tissue cells, the identification and diagnosis of complex diseases using machine learning methods has become very important. Recognizing cell characteristics such as size, shape, and chromatin design is essential in determining cell type, which can be achieved through learning methods such as deep network training. Finding the nucleus or cytoplasm of cells in medical images is a small but significant part of a long process of diagnosing and treating diseases. Today, artificial intelligence has rushed to the aid of experts in this field and has increased the speed and accuracy of experts in finding these cells and their nuclei. This... 

Recent approaches in the field of semantic image segmentation rely on deep networks that are trained by pixel-level labels. This level of labeling requires a lot of time for the labeler person; because these networks require large training datasets to achieve optimal accuracy and the lack of data at the labeled pixel level causes a significant drop in their performance. In order to overcome this problem, weakly supervised segmentation approaches have been proposed. In these approaches, weaker labels such as image-level labels, bounding boxes, scribbles, etc. have been introduced to train the networks.In this thesis, a method for segmentation of kidney and kidney tumor in CT scan images based... 

The problem of partitioning vertices of a graph has been studied with different formulations according to their application. In this thesis we try to review some of these formulations and existing algorithms. In addition we try to investigate the correct definition for a specific application in image processing, namely image segmentation. At the end, we propose a new algorithm for partitioning vertices of a weighted graph according to the mentioned application and compare its performance with some similar algorithms  

Laparoscopic surgery is a branch of minimally invasive surgery that is implemented in the abdominal cavity. This kind of surgery is conducted using surgical instruments that are inserted into the abdomen through some small incisions created on its wall and the necessary vision is provided for the surgeon using a laparoscope lens that is inserted through the first created incision. In recent years, because of its advantages, controlling and manipulation of the lens is done using some robots instead of surgeon assistant. To facilitate the controlling procedure of these robots for the surgeon, several controlling modes such as keyboard, foot switch, and voice commands are presented. Robolens is... 

Level set methods (LSM) are a conceptual framework for using level sets as a tool for numerical analysis of surfaces and shapes. The advantage of the level set model is that one can perform numerical computations involving curves and surfaces on a fixed Cartesian grid without having to parameterize these objects (this is called the Eulerian approach). Also, the level set method makes it very easy to follow shapes that change topology, for example when a shape splits in two, develops holes, or the reverse of these operations. All these make the level set method a great tool for modeling time-varying objects, like inflation of an airbag, or a drop of oil floating in water  

Image segmentation is a technique which divides an image into significant parts. The accuracy of this technique plays an important role when it applies on medical images. Among various image segmentation methods, clustering methods have been extensively investigated and used. Since it is an unsupervised method, the existence of a small amount of side-information which is extracted from a specific application (in this case, medical image) could improve its accuracy. Using this side-information in clustering methods introduces a new generation of clustering approaches called semi-supervised clustering. This information usually has a format of pair-wise constraints and can be prepared easily... 

Video segmentation means grouping of pixels of the video sequences into spatio-temporal regions which exhibit coherence in both appearance and motion. Due to complexity and spatio-temporal variations, dynamic texture segmentation is a one of the most challenging task in video processing. The problem of dynamic texture segmentation has received considerable attention due to the explosive growth of its applications in video analysis and surveillance systems. In this thesis, two novel approaches have been proposed. The first proposed method is based on generative Dynamic texture models (DTMs) which represent videos as a linear dynamical system. Since DTMs cannot be used for complex videos which... 

Scene classification and understanding is one of the most important fields in computer vision. Its applications are such as exploring robot navigation enviroment, content-based image retrieval (CBIR), organization in image databases, highly semantic describing images and videos and content extraction of videos.Many methods and algorithm are proposed till today to deal with diversity of this field by emphesizing on feature based methods or machine learning based methods. In this research we have focoused on proposing a new algorithm which is using principals of NBNN image classification method but major changes in how to exract distance metric from Nearest neighbour and how to use local... 

Digital watermarking is an efficacious technique to protect the copyright and ownership of digital information. Traditional image watermarking algorithms embed a logo in the image that could reduce its visual quality. A new approach in watermarking called zero watermarking doesn’t need to embed a logo in the image. In this algorithm we find a feature from the main image and combine it with a logo to obtain a key. This key is securely kept by a trusted authority. In this thesis we show that we can increase the robustness of digital zero watermarking by a new counter detection method in comparison to Canny Edge detection and morphological dilatation that is mostly used by related works....