Skeletonization of Digital Objects

Proceedings of the 2011 joint ACM workshop on Human gesture and behavior understanding

The article presents an analytical model of a digital information object (AObj) designed for identifying and classifying social media profiles. The model formalizes the description of digital objects through five key components: unique identifier (ID), static (S), dynamic (D), and computed (F) characteristics, along with relationships to other objects (Rel). A comprehensive methodology was developed for model construction, including: ranking characteristics by their significance for specific analytical tasks, converting ranked lists into normalized weighting coefficients (0-1 range), assigning deterministic assessments to characteristic values (numerical, scoring, or thesaurus-based), transforming to relative dimensionless values for integration into a unified criterion, and calculating integral estimates of target computed characteristics as weighted sums of relative values. Specialized methods for analyzing textual and non-textual data within the model were proposed. Specifically, base and specialized thesauri for processing text fields were created, with each thesaurus section assigned weights based on semantic load and usage frequency. For analyzing musical preferences, an audio thesaurus was developed containing compositions thematically related to adolescent protest and self-destruction. Neural network methods were employed for processing avatars and videos, enabling identification of specific visual markers (e.g., presence/absence of human faces, anonymizing elements, symbolism). Experimental results validated the proposed methods for associating accounts with the "school student" social group and detecting bots. The study developed and experimentally verified an analytical digital object model for formalized description of social media profiles. A methodology for calculating integral estimates was proposed, and specialized methods for analyzing textual, audio, and visual data were created. Testing on a sample of 800 profiles confirmed the model's effectiveness for identifying school students, bots, and at-risk users. The research outcomes have practical significance for sociology, psychology, information security, and developing intelligent social data analysis systems.

We describe a digital design process that interfaces real-time data with 3D modeling and 3D printing techniques. Digital Impressionism (DI) is a platform that explores new material possibilities, by 3D modeling physical and digital objects, as affected by invisible forces that act upon them in real time. Using a 3D pointcloud as a medium, we describe an experimental project run with students in our programme, incorporating realtime audio data to manipulate 3D physical forms, resulting in new static and dynamic shapes with what we call a hybrid materiality. The modeling platform of DI treats materials as composites, through which substance becomes physical via the digital interference the environment applies to digital forms. We describe dynamic modeling processes through which data enables a new hybrid tectonic made of composite shapes and materials. This abstract introduces the project and describes our methodology and results so far.

The GRAFIX I system was developed in the late 1960's as a fast flexible system for processing and analyzing filmed images, particularly of material which is essentially binary (black and white), such as printed text, line drawings, certain biomedical images, fingerprints, etc. It incorporates a large scale general purpose time shared computer to provide the facilities for the efficient development of algorithms necessary to perform various image processing and analysis tasks. In addition it contains a fast, high resolution flying-spot film scanner and a powerful and rather general slave processor (the binary image processor, or BIP) which provide data collection and manipulation facilities adequate to perform image processing and analysis tasks at commercially practical speeds. So far it has been successfully employed in a commercial environment to the reading of printed multifont text in complex page formats, and to the reading of Cyrillic, Greek and even handprinted text. At present we are considering future applications beyond the area of optical character recognition, particularly the analysis of engineering drawings, as well as the automatic analysis and classification of fingerprints, the analysis of biomedical images such as chromosomes, the analysis of x-ray images, and the analysis of satellite imagery data, among others.

Augmented Reality (AR) is a technology that can be used to provide personalized and contextualized information regarding physical objects in form of digital overlays. We use this technology in our research project Ambient Learning Spaces (ALS) to provide museum visitors with specific additional digital 3D information regarding the exhibits presented. With this technology, we enable museum curators to use a new form of transporting contextualized information without the need for additional physical space. However, the use of AR brings up new challenges for the creation and placement of digital contents into the museum space. In this context, we ran an anonymous survey on the use of AR in museums throughout Germany and studied responses of (N = 133) museum professionals. The results indicate that, although many museum professionals are interested in using AR technology, currently the integration is very costly and complex. This paper proposes a system we developed in a user-centered design process with a museum. This system provides an interface that helps museum professionals to cope with the complexity when placing and aligning digital 3D objects in their exhibition using mobile devices. Through this solution, visitors have the chance to experience the virtual objects spatially embedded in the exhibition by the curators themselves. In multiple user studies during the development phase we measured the usability of the interface. The findings show that the system provides a high degree of usability and can be applied effectively by museum professionals.

Matrix newton interpolation and progressive 3D imaging: PC-based computation

This study aims to simulate the 3 dimensional (3D) model of Keum-man connection canal using geographic information system (GIS) as well as considering design in viewpoint of engineering. The canal connects from Keumkang to Mangyungkang in order to supply fresh water into Saemankeum lake. This study used 3 dimensional spatial planning model (3DSPLAM) process to generate the 3D model, which has not only several planning layers in actual process, but also their corresponding layers in modeling process to simulate 3D space of rural villages. The discharge of the canal is <TEX>$20m^3/s$</TEX> on slope of 1/28,400 in the canal length of 14.2km, which consists of pipe line and open channel. This study surveyed the route of the canal and its surrounding environment for facilities to make images in the 3D graphic model. Besides, the present study developed data set in GIS for geogrphical surface modeling as well as parameters in hydraulic analysis for water surface profile on the canal using HEC-RAS model. From the data set constructed, this study performed analysis of water surface profile with HEC-RAS, generation of digital elevation model (DEM) and 3D objects, design of the canal section and route on DEM in AutoCAD, and 3D canal model and its surrounding 3D space in 3DMAX with virtual reality. The study result showed that the process making 3D canal model tried in this study is very useful to generate computer graphic model with the designed canal on the surface of DEM. The generated 3D canal can be used to assist decision support for the canal policy.

The work is devoted to developing the main results in solving the fundamental problem of formalization and systematization of methods and forms of information representation in image analysis, recognition, and understanding tasks. This problem arises in connection with development of the descriptive approach to image analysis and understanding. The main result is the concept of image spatial formalization, including a set of image states and a set of states of image transformation schemes. We consider the construction of algorithmic schemes generating phase trajectories for solving problems of image analysis and recognition. We present axioms defining the properties and structure of the image formalization space. The introduced system of concepts constitute the basis for standardizing methods for synthesizing image models oriented to image analysis and understanding tasks.

Deep Learning in Medical Ultrasound Analysis: A Review

Visual information, in the form of images and video, comes from the interaction of light with objects. Illumination is a fundamental element of visual information. Detecting and interpreting illumination effects is part of our everyday life visual experience. Shading for instance allows us to perceive the three-dimensional nature of objects. Shadows are particularly salient cues for inferring depth information. However, we do not make any conscious or unconscious effort to avoid them as if they were an obstacle when we walk around. Moreover, when humans are asked to describe a picture, they generally omit the presence of illumination effects, such as shadows, shading, and highlights, to give a list of objects and their relative position in the scene. Processing visual information in a way that is close to what the human visual system does, thus being aware of illumination effects, represents a challenging task for computer vision systems. Illumination phenomena interfere in fact with fundamental tasks in image analysis and interpretation applications, such as object extraction and description. On the other hand, illumination conditions are an important element to be considered when creating new and richer visual content that combines objects from different sources, both natural and synthetic. When taken into account, illumination effects can play an important role in achieving realism. Among illumination effects, shadows are often integral part of natural scenes and one of the elements contributing to naturalness of synthetic scenes. In this thesis, the problem of extracting shadows from digital images is discussed. A new analysis method for the segmentation of cast shadows in still and moving images without the need of human supervision is proposed. The problem of separating moving cast shadows from moving objects in image sequences is particularly relevant for an always wider range of applications, ranging from video analysis to video coding, and from video manipulation to interactive environments. Therefore, particular attention has been dedicated to the segmentation of shadows in video. The validity of the proposed approach is however also demonstrated through its application to the detection of cast shadows in still color images. Shadows are a difficult phenomenon to model. Their appearance changes with changes in the appearance of the surface they are cast upon. It is therefore important to exploit multiple constraints derived from the analysis of the spectral, geometric and temporal properties of shadows to develop effective techniques for their extraction. The proposed method combines an analysis of color information and of photometric invariant features to a spatio-temporal verification process. With regards to the use of color information for shadow analysis, a complete picture of the existing solutions is provided, which points out the fundamental assumptions, the adopted color models and the link with research problems such as computational color constancy and color invariance. The proposed spatial verification does not make any assumption about scene geometry nor about object shape. The temporal analysis is based on a novel shadow tracking technique. On the basis of the tracking results, a temporal reliability estimation of shadows is proposed which allows to discard shadows which do not present time coherence. The proposed approach is general and can be applied to a wide class of applications and input data. The proposed cast shadow segmentation method has been evaluated on a number of different video data representing indoor and outdoor real-world environments. The obtained results have confirmed the validity of the approach, in particular its ability to deal with different types of content and its robustness to different physically important independent variables, and have demonstrated the improvement with respect to the state of the art. Examples of application of the proposed shadow segmentation tool to the enhancement of video object segmentation, tracking and description operations, and to video composition, have demonstrated the advantages of a shadow-aware video processing.

Segmentation is always a difficult task in image analysis. In this paper, we propose a solution to computing distance transforms in images with fuzzy object borders. The difference from a standard distance transform is in the initialisation, which takes the fuzziness of the border into account. As an example of its usefulness, the new fuzzy border distance transform is used in skeletonization.

Road extraction from high-resolution remote sensing images (HRSIs) is one of the tasks in image analysis. Deep convolutional neural networks have become the primary method for road extraction due to their powerful feature representation capability. However, roads are often obscured by vegetation, buildings, and shadows in HRSIs, resulting in incomplete and discontinuous road extraction results. To address this issue, we propose a lightweight post-processing network called RIRNet in this study, which include an information inference module and a road direction inference task branch. The information inference module can infer spatial information relationships between different rows or columns of feature images from different directions, effectively inferring and repairing road fractures. The road direction inference task branch performs the road direction prediction task, which can constrain and promote the road extraction task, thereby indirectly enhancing the inference ability of the post-processing model and realizing the optimization of the initial road extraction results. Experimental results demonstrate that the RIRNet model can achieve an excellent post-processing effect, which is manifested in the effective repair of broken road segments, as well as the handling of errors such as omission, misclassification, and noise, proving the effectiveness and generalization of the model in post-processing optimization.

Though image segmentation is a fundamental task in image analysis; it plays a vital role in the area of image processing. Its value increases in case of medical diagnostics through medical images like X-ray, PET, CT and MRI. In this paper, an attempt is taken to analyse a CT brain image. It has been segmented for a particular patch in the brain CT image that may be one of the tumours in the brain. The purpose of segmentation is to partition an image into meaningful regions with respect to a particular application. Image segmentation is a method of separating the image from the background, read the contents and isolating it. In this paper both the concept of clustering and thresholding technique with edge based segmentation methods like sobel, prewitt edge detectors is applied. Then the result is optimized using GA for efficient minimization of the objective function and for improved classification of clusters. Further the segmented result is passed through a Gaussian filter to obtain a smoothed image.

An automatic quality assessment of stitched images is an essential task in image analysis and is particularly useful not only in the creation of general-purpose panoramic images but also in terrain exploration and mapping made by mobile robots and drones. In Visual Simultaneous Localization and Mapping (VSLAM) solutions, the environment maps acquired by cameras mounted on the mobile robots may be captured in dynamically changing lighting conditions and subject to some other distortions influencing the final quality of the panoramic images representing the robot’s surroundings. Such images may also be used for motion planning and visual navigation for other robots, e.g., in follow-the-leader scenarios. Another relevant application area of panoramic imaging is Virtual Reality (VR), particularly head-mounted displays, where perceived image quality is even more important. Hence, automatic quality evaluations of stitched images should be made using algorithms that are both sensitive to various types of distortions and strongly consistent with subjective quality impression. The approach presented in this paper extends the state-of-the-art metric known as the Stitched Image Quality Evaluator (SIQE) by embedding it with some additional features using the proposed new combination scheme. The developed combined metric based on a nonlinear combination of the SIQE and additional features led to a substantially higher correlation with the subjective quality scores.

An important task in image analysis is to assign each pixel in an image to a predefined class of objects. This class can be a specific anatomy like the liver in a computed tomography image. One commonly used step is pixel classification where each pixel is assigned a class-based purely on its value and independently of the pixel neighbors. In this chapter, we introduce basic pixel classification approaches like minimum distance classification and parametric classification, where the pixel value distributions are approximated using Gaussian kernels. Finally, a brief overview of Bayesian classification is given.