 |
|
Human eye inspires advance in computer vision from Boston College researchers
June 18, 2009Inspired by the behavior of the human eye, Boston College computer scientists have developed a technique that lets computers see objects as fleeting as a butterfly or tropical fish with nearly double the accuracy and 10 times the speed of earlier methods.
The linear solution to one of the most vexing challenges to advancing computer vision has direct applications in the fields of action and object recognition, surveillance, wide-base stereo microscopy and three-dimensional shape reconstruction, according to the researchers, who will report on their advance at the upcoming annual IEEE meeting on computer vision.
BC computer scientists Hao Jiang and Stella X. Yu developed a novel solution of linear algorithms to streamline the computer's work. Previously, computer visualization relied on software that captured the live image then hunted through millions of possible object configurations to find a match. Further compounding the challenge, even more images needed to be searched as objects moved, altering scale and orientation.
Rather than combing through the image bank - a time- and memory-consuming computing task - Jiang and Yu turned to the mechanics of the human eye to give computers better vision.
"When the human eye searches for an object it looks globally for the rough location, size and orientation of the object. Then it zeros in on the details," said Jiang, an assistant professor of computer science. "Our method behaves in a similar fashion, using a linear approximation to explore the search space globally and quickly; then it works to identify the moving object by frequently updating trust search regions."
Trust search regions act as visual touchstones the computer returns to again and again. Jiang and Yu's solution focuses on the mathematically-generated template of an image, which looks like a constellation when lines are drawn to connect the stars. Using the researchers' new algorithms, computer software identifies an object using the template of a trust search region. The program then adjusts the trust search regions as the object moves and finds its mathematical matches, relaying that shifting image to a memory bank or a computer screen to record or display the object.
Jiang says using linear approximation in a sequence of trust regions enables the new program to maintain spatial consistency as an object moves and reduces the number of variables that need to be optimized from several million to just a few hundred. That increased the speed of image matching 10 times over compared with previous methods, he said.
The researchers tested the software on a variety of images and videos - from a butterfly to a stuffed Teddy Bear - and report achieving a 95 percent detection rate at a fraction of the complexity. Previous so-called "greedy" methods of search and match achieved a detection rate of approximately 50 percent, Jiang said.
Boston College
Rather than combing through the image bank - a time- and memory-consuming computing task - Jiang and Yu turned to the mechanics of the human eye to give computers better vision.
"When the human eye searches for an object it looks globally for the rough location, size and orientation of the object. Then it zeros in on the details," said Jiang, an assistant professor of computer science. "Our method behaves in a similar fashion, using a linear approximation to explore the search space globally and quickly; then it works to identify the moving object by frequently updating trust search regions."
Trust search regions act as visual touchstones the computer returns to again and again. Jiang and Yu's solution focuses on the mathematically-generated template of an image, which looks like a constellation when lines are drawn to connect the stars. Using the researchers' new algorithms, computer software identifies an object using the template of a trust search region. The program then adjusts the trust search regions as the object moves and finds its mathematical matches, relaying that shifting image to a memory bank or a computer screen to record or display the object.
Jiang says using linear approximation in a sequence of trust regions enables the new program to maintain spatial consistency as an object moves and reduces the number of variables that need to be optimized from several million to just a few hundred. That increased the speed of image matching 10 times over compared with previous methods, he said.
The researchers tested the software on a variety of images and videos - from a butterfly to a stuffed Teddy Bear - and report achieving a 95 percent detection rate at a fraction of the complexity. Previous so-called "greedy" methods of search and match achieved a detection rate of approximately 50 percent, Jiang said.
Boston College
Computer Vision Current Events and Computer Vision News RSSMU Research Leads to Improved Human, Object Detection Technology
When searching for basketball videos online, a long list of websites appears, which may contain a picture or a word describing a basketball. But what if the computer could search inside videos for a basketball?
Scientists develop an 'intelligent car' able to learn from his owner's driving and warn him in case of accident hazard
UGR News Scientists from six European countries, including Spain, have developed a new computer system so called DRIVSCO that allows vehicles to learn from the behaviour of their drivers at the wheel, in such a way that they can detect if a driver presents an "unusual behaviour" in a curve or an obstacle on the road and generates signals of alarm which warn him on time to react.
Rome was built in a day, with hundreds of thousands of digital photos
The ancient city of Rome was not built in a day. It took nearly a decade to build the Colosseum, and almost a century to construct St. Peter's Basilica. But now the city, including these landmarks, can be digitized in just a matter of hours.
Successful neurosurgery with transcranial MR-guided high-intensity focused ultrasound
The Magnetic Resonance Center of the University Children's Hospital Zurich has achieved a world first break through in MR-guided, non-invasive neurosurgery.
MIT: Why we have difficulty recognizing faces in photo negatives
Humans excel at recognizing faces, but how we do this has been an abiding mystery in neuroscience and psychology. In an effort to explain our success in this area, researchers are taking a closer look at how and why we fail.
More effective treatment identified for common childhood vision disorder
Scientists have found a more effective treatment for a common childhood eye muscle coordination problem called convergence insufficiency (CI).
New robotic repair system will fix ailing satellites
Researchers at Queen's University are developing a new robotic system to service more than 8,000 satellites now orbiting the Earth, beyond the flight range of ground-based repair operations.
Passports for penguins
Ground-breaking technology that will enable biologists to identify and monitor large numbers of endangered animals, from butterflies to whales, without being captured, will be shown to the public for the first time at this year's Royal Society Summer Science exhibition [30 June to 3 July].
UC San Diego computer scientist turns his face into a remote control
A computer science Ph.D. student can turn his face into a remote control that speeds and slows video playback. The proof-of-concept demonstration is part of a larger project to use automated facial expression recognition to make robots more effective teachers.
Carnegie Mellon system estimates geographic location of photos
Researchers at Carnegie Mellon University have devised the first computerized method that can analyze a single photograph and determine where in the world the image likely was taken. It's a feat made possible by searching through millions of GPS-tagged images in the Flickr online photo collection.
More Computer Vision Current Events and Computer Vision News Articles
 |
Computer Vision by Linda G. Shapiro (Author), George C. Stockman (Author) (Pearson Education) A textbook and reference for students and practitioners, presenting the necessary theory for work in fields where significant information must be extracted from images. Topics covered include databases and virtual and augmented reality, and the text includes more than 250 exercises and programming projects. DLC: Computer vision. |
 |
Machine Vision, Third Edition: Theory, Algorithms, Practicalities (Signal Processing and its Applications) by E. R. Davies (Author) In the last 40 years, machine vision has evolved into a mature field embracing a wide range of applications including surveillance, automated inspection, robot assembly, vehicle guidance, traffic monitoring and control, signature verification, biometric measurement, and analysis of remotely sensed images. While researchers and industry specialists continue to document their work in this area, it has become increasingly difficult for professionals and graduate students to understand the essential theory and practicalities well enough to design their own algorithms and systems. This book directly addresses this need. As in earlier editions, E.R. Davies clearly and systematically presents the basic concepts of the field in highly accessible prose and images, covering essential... |
 |
Learning OpenCV: Computer Vision with the OpenCV Library by Gary Bradski (Author), Adrian Kaehler (Author), Bradski Gary (Author), Kaehler Adrian (Author) "This library is useful for practitioners, and is an excellent tool for those entering the field: it is a set of computer vision algorithms that work as advertised." -William T. Freeman, Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology Learning OpenCV puts you in the middle of the rapidly expanding field of computer vision. Written by the creators of the free open source OpenCV library, this book introduces you to computer vision and demonstrates how you can quickly build applications that enable computers to "see" and make decisions based on that data. Computer vision is everywhere-in security systems, manufacturing inspection systems, medical image analysis, Unmanned Aerial Vehicles, and more. It stitches Google maps... |
 |
Computer Vision: A Modern Approach by David A. Forsyth (Author), Jean Ponce (Author) The accessible presentation of this book gives both a general view of the entire computer vision enterprise and also offers sufficient detail to be able to build useful applications. Users learn techniques that have proven to be useful by first-hand experience and a wide range of mathematical methods. A CD-ROM with every copy of the text contains source code for programming practice, color images, and illustrative movies. Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance. Topics are discussed in substantial and increasing depth. Application surveys describe numerous important application areas such as image based rendering and digital libraries. Many important algorithms broken down and... |
 |
Multiple View Geometry in Computer Vision by Richard Hartley (Author), Andrew Zisserman (Author) A basic problem in computer vision is to understand the structure of a real world scene. This book covers relevant geometric principles and how to represent objects algebraically so they can be computed and applied. Recent major developments in the theory and practice of scene reconstruction are described in detail in a unified framework. Richard Hartley and Andrew Zisserman provide comprehensive background material and explain how to apply the methods and implement the algorithms. First Edition HB (2000): 0-521-62304-9 |
 |
An Introduction to 3D Computer Vision Techniques and Algorithms by Boguslaw Cyganek (Author), J. Paul Siebert (Author) Computer vision encompasses the construction of integrated vision systems and the application of vision to problems of real-world importance. The process of creating 3D models is still rather difficult, requiring mechanical measurement of the camera positions or manual alignment of partial 3D views of a scene. However using algorithms, it is possible to take a collection of stereo-pair images of a scene and then automatically produce a photo-realistic, geometrically accurate digital 3D model.
This book provides a comprehensive introduction to the methods, theories and algorithms of 3D computer vision. Almost every theoretical issue is underpinned with practical implementation or a working algorithm using pseudo-code and complete code written in C++ and MatLab®. There is... |
 |
Machine Vision Algorithms and Applications by Carsten Steger (Author), Markus Ulrich (Author), Christian Wiedemann (Author) This first up-to-date textbook for machine vision software provides all the details on the theory and practical use of the relevant algorithms. The first part covers image acquisition, including illumination, lenses, cameras, frame grabbers, and bus systems, while the second deals with the algorithms themselves. This includes data structures, image enhancement and transformations, segmentation, feature extraction, morphology, template matching, stereo reconstruction, and camera calibration. The final part concentrates on applications, and features real-world examples, example code with HALCON, and further exercises. Uniting the latest research results with an industrial approach, this textbook is ideal for students of electrical engineering, physics and informatics, electrical and... |
 |
Algorithms for Image Processing and Computer Vision by J. R. Parker (Author) A cookbook of the hottest new algorithms and cutting-edge techniques in image processing and computer vision This amazing book/CD package puts the power of all the hottest new image processing techniques and algorithms in your hands. Based on J. R. Parker's exhaustive survey of Internet newsgroups worldwide, Algorithms for Image Processing and Computer Vision answers the most frequently asked questions with practical solutions. Parker uses dozens of real-life examples taken from fields such as robotics, space exploration, forensic analysis, cartography, and medical diagnostics, to clearly describe the latest techniques for morphing, advanced edge detection, wavelets, texture classification, image restoration, symbol recognition, and genetic algorithms, to... |
 |
Introductory Techniques for 3-D Computer Vision by Trucco (Author), Alessandro Verri (Author) Senior/Graduate level courses on computer vision, robot vision and image processing in electrical and computer engineering, mathematics, and computer science departments, and an essential reference for researchers and scientists in the field of computer vision. An applied introduction to modern computer vision, focusing on a set of computational techniques for 3-D imaging. Covers a wide range of fundamental problems encountered within computer vision and provides detailed algorithmic and theoretical solutions for each. Each chapter concentrates on a specific problem and solves it by building on previous results. |
 |
3D Computer Vision: Efficient Methods and Applications (X.media.publishing) by Christian Wöhler (Author) This book provides an introduction to the foundations of three-dimensional computer vision and describes recent contributions to the field. Geometric methods include linear and bundle adjustment based approaches to scene reconstruction and camera calibration, stereo vision, point cloud segmentation, and pose estimation of rigid, articulated, and flexible objects. Photometric techniques evaluate the intensity distribution in the image to infer three-dimensional scene structure, while real-aperture approaches exploit the behavior of the point spread function. It is shown how the integration of several methods increases reconstruction accuracy and robustness. Applications scenarios include industrial quality inspection, metrology, human-robot-interaction, and remote sensing. |
| © 2009 BrightSurf.com |