Zamansiz

Intel Atom™ Developer Program Million Dollar Development Fund

Zamansiz — Wed, 03 Mar 2010 06:20:43 +0000

The netbook has become one of the most popular consumer devices in the market today, but its true potential has been limited by applications that are not optimized for its mobility and smaller screen size.

The Intel Atom Developer Program Million Dollar Development Fund is a fund set up by Intel to address this challenge and to help accelerate innovation in software applications for Intel? Atom™ processor-based products, starting with netbooks, and eventually supporting smartphones, consumer electronics and more devices.

The fund will be available to individual and student developers as well as small, medium and large software companies to support development of groundbreaking applications for the netbook platform. The first three elements of the development fund Fast Track 2010, Dollars for Downloads 2010 and the Intel Atom Developer Challenge are available now with more to be announced over the coming months.

Eric Giler demos wireless electricity

Zamansiz — Tue, 09 Feb 2010 09:44:06 +0000

WiTricity Technology: WiTricity power sources and capture devices are specially designed magnetic resonators that efficiently transfer power over large distances via the magnetic near-field. These proprietary source and device designs and the electronic systems that control them support efficient energy transfer over distances that are many times the size of the sources/devices themselves.

The WiTricity power source, left, is connected to AC power. The blue lines represent the magnetic near field induced by the power source. The yellow lines represent the flow of energy from the source to the WiTicity capture coil, which is shown powering a light bulb. Note that this diagram also shows how the magnetic field (blue lines) can wrap around a conductive obstacle between the power source and the capture device.

Online Lecture : Recognizing and Learning Object Categories

Zamansiz — Sun, 07 Feb 2010 14:42:51 +0000

ICCV 2009 Recognizing and Learning Object Categories: Year 2009

Introduction (.pptx, .pdf)
Part 1: Single object classes
- Bag of words models, Part-based models, and Discriminative models (.pptx)
- Detecting single objects in context (.pptx)
- 3D object models (.pptx)
Part 2: Multiple object categories
- Recognizing a large number of object classes (.pptx)
- Recognizing multiple objects in an image. Sharing and context (.pptx)
- Objects and annotations (.pptx)
Part 4: Summary and datasets (.pptx)

for more info please visit http://people.csail.mit.edu/torralba/shortCourseRLOC/index.html

RAMdESC – Real-time automatic object recognition

Zamansiz — Thu, 04 Feb 2010 15:21:28 +0000

The goal of the projects is to develop new algorithms for medical image analysis and visualization of medical imagery, as well as to build vision based systems for surgical navigation and surgical planning.

The project team page is will be avaible online 22 March on Zamansiz.org

What is Continuous Integration

Zamansiz — Wed, 27 Jan 2010 11:15:02 +0000

The practice of continuous integration represents a fundamental shift in the process of building software. It takes integration, commonly an infrequent and painful exercise, and makes it a simple, core part of a developer’s daily activities. Integrating continuously makes integration a part of the natural rhythm of coding, an integral part of the test-code-refactor cycle. Continuous integration is about progressing steadily forward by taking small steps.

Integration should happen continuously, and continuously is more often than you might think. The frequency of integration will vary from project to project, from developer to developer, and from modification to modification. However, as a goal and a good rule of thumb, developers should integrate their changes once every few hours and at least once per day.

Learning how to integrate so frequently requires practice and discipline. Fundamentally, an integration can occur at any point when the code compiles and all the unit tests are passing. The challenge is learning how to write software so that you never stray too far from this point. If you are testing at the right level of granularity and are refactoring regularly, then you should never be more than a few minutes away from this point. This means that you are almost always in a position where you can launch a new integration.

Deciding when to integrate is all about controlling risk. When making modifications in a high traffic area of the code base or when conducting broad refactorings like class renaming or package reorganisation, there is an elevated risk of impacting other developers or of having merge conflicts when committing. The longer that developers go without integrating, the greater the likelihood of conflicts and the larger the effort required to resolve those conflicts. As the effort of integration increases exponentially in proportion to the time between integrations, best practices dictate that when making high-risk changes a developer should start from a clean workspace, focus only on required modifications, proceed with the smallest logical steps, and then commit at the earliest opportunity.

A successful integration is a measure of progress. It provides feedback that the new code runs correctly in the integration environment and successfully interoperates with the rest of the code base. Code sitting unintegrated in a developer’s workspace simply does not exist. It is not part of the code base, it cannot be accessed by other developers or tested by the customer. Only when it has been successfully integrated is the benefit of the new code realised.

References

This text is snipped from Owen Rogers’ recent article: Scaling Continuous Integration submitted to XP2004
Martin Fowler and Matt Foemmel’s classic article: Continuous Integration

Video Object Segmentation

Zamansiz — Sat, 23 Jan 2010 09:44:27 +0000

A few days ago I found a very good thesis about video object segmentation.You have to read it carefully if you want to understand basic and advanced things of computer vision. thank you Fatih Murat PORIKLI for excellent work

here you can download the fulltext

Download thesis (low-res, 7.5Mb)

Excerpt from the introduction:
More and more visual information is available in digital form, in various places and on various media. The emergence of digital video and its proliferation in multimedia applications has created a significant demand for content-based representation of visual information. Main purpose of video segmentation is to enable content-based representation by extracting objects of interest from a series of consecutive video frames. Briefly, the motivation behind video segmentation can be categorized as the applications in indexing and retrieval, compression and coding, recognition, identification, and understanding of video scenes, editing, manipulation, and animation.

Video databases on the market today allow only limited capability of or domain limited searching for video using characteristics like color, texture, and simpler motion statistics. If video can be stored in the form of individual objects, indexing and retrieval of visual information is as simple as that of textual information. An essential tool in the management of visual records is the ability to automatically describe and index the content of video sequences in a meaningful manner. Such a facility would allow recovery of desired video segments or objects from a very large database of video sequences. The efficient use of stock film archives and identification of specific activities in surveillance videos are among the potential applications.

From a compression point of view, video segmentation is essential for object-based video coding standards, i.e. MPEG-4. Due to the vast data size of video sequences, communicating digital video over the bandwidth limited network sources demands competent coding techniques. Having an object-based representation scheme that identifies the important parts of image frames, video sequences can also be encoded efficiently to satisfy transmission requirements. Videoconferencing is one of the applications that benefit from object-based coding.

Video segmentation is key to many robotic vision applications. Most vision based autonomous vehicles acquire information on their surroundings by analyzing video. Particularly, it is required for high-level image understanding and scene interpretation such as spotting and tracking of special events in surveillance video. For instance, pedestrian and highway traffic can be regularized using density evaluations obtained by segmenting people and vehicles. By object segmentation, speeding and suspicious moving cars, road obstacles, strange activities can be detected. Forbidden zones, parking lots, elevators can be monitored automatically. Gesture recognition as well as visual biometric extraction can be done for user interfaces.

With a good segmentation, it is possible to access and manipulate objects in video. To illustrate, traffic enforcement currently employs supervised video segmentation tools to acquire identity of speeding or trespassing cars. Infotainment industry utilizes video segmentation for editing, manipulating, and animation.

Although the human being can quickly interpret the embedded semantic content from the information carried by different modalities, computer understanding of visual information is still in its primitive stage. Good segmentation tools are crucial to the success of the future standards. But tasks of automatically segmenting image sequences into semantic meaningful objects prove to be very challenging. We have currently a reasonably good understanding of the basic mechanisms underlying visual information processing, still, many questions are still open to investigation, some desperately waiting for an answer.

Contents

1 Introduction 1
1.1 Motivation of Video Segmentation . . . . . . . . . . . . . . . . . . . . 2
1.2 Object: A Bridge from Pixels to Semantic . . . . . . . . . . . . . . . 5
1.3 Elementary Categorization . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Video Coding Standards . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Scope of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.6 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Background on Video Segmentation 12
2.1 Region Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.1 Histogram Thresholding in Color Space . . . . . . . . . . . . . 16
2.1.2 Clustering in Color Space . . . . . . . . . . . . . . . . . . . . 19
2.1.3 Region Growing . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1.4 Morphological and Edge Based Techniques . . . . . . . . . . . 26
2.1.5 Split and Merge Techniques . . . . . . . . . . . . . . . . . . . 27
2.1.6 Texture Segmentation . . . . . . . . . . . . . . . . . . . . . . 31
2.2 Motion Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.1 Block-Matching . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.2 Feature Point Matching . . . . . . . . . . . . . . . . . . . . . 36
2.2.3 Optical Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.2.4 Nonparametric and Parametric Motion Models . . . . . . . . . 39
2.3 Spatio-Temporal Segmentation . . . . . . . . . . . . . . . . . . . . . . 40
2.3.1 Change Detection Mask . . . . . . . . . . . . . . . . . . . . . 41
2.3.2 Stochastic Approaches . . . . . . . . . . . . . . . . . . . . . . 43
2.3.3 Morphological Approaches . . . . . . . . . . . . . . . . . . . . 45
2.3.4 Hybrid Approaches . . . . . . . . . . . . . . . . . . . . . . . . 46
2.4 Object Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.4.1 Object Models for Tracking . . . . . . . . . . . . . . . . . . . 50
2.4.2 Tracking Techniques . . . . . . . . . . . . . . . . . . . . . . . 53
2.5 Segmentation in Compressed Domain . . . . . . . . . . . . . . . . . . 55
2.6 Scene-Cut Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.7 Data Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3 Preprocessing of Color Digital Video for Segmentation 63
3.1 Analysis of Suitable Attributes . . . . . . . . . . . . . . . . . . . . . . 64
3.1.1 Color Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.1.2 Comparison of Color Spaces . . . . . . . . . . . . . . . . . . . 72
3.1.3 Texture Elements . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.1.4 Edge Elements . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.2 Filtering and Simplification . . . . . . . . . . . . . . . . . . . . . . . 81
3.2.1 Implementation of Fast Median Filter . . . . . . . . . . . . . . 83
3.2.2 Low-Pass Filtering by Gaussian Kernels . . . . . . . . . . . . 85
3.2.3 Smoothing by Morphological Operators . . . . . . . . . . . . . 86
3.2.4 Image Simplification by Robust Estimators . . . . . . . . . . . 88
3.2.5 Recursive Band-Suppression Filters . . . . . . . . . . . . . . . 94
3.2.6 Comparison of Filters . . . . . . . . . . . . . . . . . . . . . . . 98
3.3 Change Detection Mask . . . . . . . . . . . . . . . . . . . . . . . . . 100
3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4 An Unsupervised Moving Object Segmentation Framework 106
4.1 System Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
4.2 Formation of Spatiotemporal Data Structure . . . . . . . . . . . . . . 111
4.2.1 Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.2.2 Color Quantization & MPEG-7 . . . . . . . . . . . . . . . . . 115
4.3 Marker Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4.3.1 Uniformly Distributed Markers . . . . . . . . . . . . . . . . . 119
4.3.2 Minimum Gradient Points as Markers . . . . . . . . . . . . . . 120
4.4 Volume Growing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
4.4.1 Linkage Methods of Volume Growing . . . . . . . . . . . . . . 123
4.4.2 Single-linkage Algorithm . . . . . . . . . . . . . . . . . . . . . 124
4.4.3 Centroid-linkage Algorithm . . . . . . . . . . . . . . . . . . . 126
4.4.4 Dual-linkage Algorithm . . . . . . . . . . . . . . . . . . . . . . 129
4.4.5 Threshold Determination . . . . . . . . . . . . . . . . . . . . . 131
4.4.6 Modes of Volume Growing . . . . . . . . . . . . . . . . . . . . 134
4.4.7 Volume Refinement . . . . . . . . . . . . . . . . . . . . . . . . 136
4.5 Analysis of Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.5.1 Extraction of Trajectories . . . . . . . . . . . . . . . . . . . . 140
4.5.2 Quantitative Descriptors . . . . . . . . . . . . . . . . . . . . . 144
4.5.3 Relational Descriptors . . . . . . . . . . . . . . . . . . . . . . 145
4.5.4 Change Detection Mask in Segmentation . . . . . . . . . . . . 148
4.5.5 Color Detection Mask . . . . . . . . . . . . . . . . . . . . . . 149
4.5.6 Feature-based Motion Estimation . . . . . . . . . . . . . . . . 151
4.6 Clustering Volumes into Objects . . . . . . . . . . . . . . . . . . . . . 155
4.6.1 Fine-to-Coarse Hierarchy . . . . . . . . . . . . . . . . . . . . . 155
4.6.2 Coarse-to-Fine Hierarchy . . . . . . . . . . . . . . . . . . . . . 160
4.7 Multi-Resolution Object Tree . . . . . . . . . . . . . . . . . . . . . . 163
4.8 Test Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
4.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
5 Concluding Remarks 172
5.1 Summary of Main Contributions . . . . . . . . . . . . . . . . . . . . . 172
5.2 Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

8th European Machine Vision Business Conference, April 16th & April 17th, 2010

Zamansiz — Sat, 08 Aug 2009 18:18:50 +0000

Please mark your calendar!

The 8th EMVA Business Conference 2010 will take place on April 16th & April 17th, 2010 in Istanbul / Turkey.

Further information will follow within the next weeks.

Please mark your calendar!

The 8th EMVA Business Conference 2010 will take place on April 16th & April 17th, 2010 in Istanbul / Turkey.

Further information will follow within the next weeks.

Location

Istanbul

Turkey

For more information please visit www.emva.org

National Center for High Performance Computing

Zamansiz — Fri, 24 Jul 2009 08:35:59 +0000

Magnificent improvements on computation power, memory and data transfer technologies made the computational scienses and engineering revolutionary grow. In most application domains, high grade results could be achieved by using only computation. A few of the research fields are using computation are climate researchs, earth science, nano-technologies, computational chemistry, high energy physics, computational fluid mechanics, life sciences etc.

The main goals of the National Center for High Performance Computing are to build awareness regarding to computational sciences and engineering in Turkey and to make ready a computational infra-structure for scientific researches and R&D services.

National Center for High Performance Computing Project started in 2004 with the support of DPT [Devlet Planlama Teskilati(Prime Ministry State Planning Organization)].

Three different user groups are targeted in our HPC Center:

Scientific researchers that are made in universities and public sector’s research departments,
R&D departments of industrial companies that need computational resources for their services,
The projects of international research and application.

In November 2006, the first phase of the server system is deployed in its temporary location in Istanbul Technical University Ayazaga Campus. Deployment process is finished in January 2007 and it was opened for users.

The second phase of the server system, disk and tape based storage system and virtual reality laboratuaries will be in use after the permanent building is finished.

for more information please visit : http://www.uybhm.itu.edu.tr/eng/index.html

GPU Technology Conference

Zamansiz — Wed, 15 Jul 2009 19:05:33 +0000

September 30-October 2, 2009 – The Fairmont Hotel, San Jose, California

The GPU Developers Summit is designed to help developers of consumer, professional, and HPC applications to harness the massively parallel processing power of the GPU. Experts from a broad range of industries will share insights and updates on state of the art techniques in GPU Computing, media processing, advanced visualization and related areas. Presentations will cover a wide range of topics using industry standard languages such as C/C++ with CUDA extensions and Fortran, GPU Computing APIs such as DirectX Compute and OpenCL™ on the CUDA Architecture, graphics API’s such as Direct3D and OpenGL and powerful libraries and middleware.

Topic clusters, subject to change, will cover both computation and graphics, spanning across a broad range of industries and interests, in research and commercial applications, such as

Computer vision	Digital content creation	C/C++ with CUDA Extensions
Augmented reality	Video and image processing	DirectX Compute on the GPU
Visual analytics	Advanced visualization	OpenCL on the GPU
Supercomputing	Computer aided engineering	OpenGL
Computational finance	Environmental simulation	DirectX 11
Computational fluid dynamics	Reality simulation	Medical imaging
Advanced numeric computing	3D Stereo	Life sciences
GPU computing in education	Scientific visualization	Energy exploration
AstroGPU, BioGPU, X on GPU	Visual simulation	Film and broadcast
MultiGPU		Automotive

Sessions topics, along with a full conference catalog, will be published in the coming weeks. Stay up-to-date by subscribing to our email updates.

If you are interested in submitting a proposal for tutorials, sessions, panels, birds of a feather, posters, or roundtables, please see our Call for Submissions page.

Maximize Your Conference Experience with Pre-Event Tutorials and Webinars
Take advantage of our pre-conference tutorials and webinars to get up to speed on programming languages and APIs for the GPU. Doing so prior to the event will help you maximize your time during the advanced technical sessions and discussions. Pre-conference tutorials will be held before the keynotes on Wednesday, September 30 (please see schedule for details). Additionally, there are on-going webinars and other archived resources on NVIDIA.com and CUDA Zone.

ISC 2009 CUDA/OpenCL Tutorial Slides Posted

Zamansiz — Fri, 03 Jul 2009 12:26:09 +0000

A tutorial on High Performance Computing with CUDA was held at the International Conference on Supercomputing in Hamburg on Monday, June 22nd 2009. The tutorial included an introduction to the CUDA programming model and C for CUDA, along with details on the CUDA Toolkit, Libraries, and optimization. The tutorial also provided an introduction to OpenCL, and finished with a case study on Computational Fluid Dynamics by Dr. Graham Pullan from Cambridge University. Slides from the tutorial are now posted here on GPGPU.org.

(Massimiliano Fatica, Timo Stich, and Graham Pullan. High Performance Computing with CUDA. Tutorial. International Conference on Supercomputing 2009. Hamburg, Germany.)