Decoupled Active Contours – VISION AND IMAGE PROCESSING (VIP) RESEARCH GROUP

Akshaya Mishra

vipadmin — Thu, 23 Mar 2023 18:04:04 +0000

Akshaya has been working in the field of image processing, pattern recognition and computer vision for nine years. He has been publishing and serving as an reviewer for several key image processing and computer vision conferences and journals.

He has obtained a Master of Technology in Automation and Computer Vision from Indian Institute of Technology, Kharagpur, India and a Ph.D. degree in pattern analysis and machine intelligence from Systems Design Department of University of Waterloo. Prior to starting his Ph.D, Akshaya worked for two years and six months for Read-Ink Technology Pvt. Ltd to develop machine learning and pattern recognition algorithms for online hand written character recognition engine.

After completing his Ph.D. Akshaya worked for one year for Tornado Medical Systems, where he developed algorithms for reconstruction and analysis of Optical Coherence Tomography (OCT) signals. Recently, Akshaya has joined Miovision Technologies Pvt. Ltd to work in the areas of intelligent traffic data analysis and management. Akshaya’s interest includes but are not limited to document image analysis, medical image reconstruction and analysis with a focus on Optical Coherence Tomography signals, traffic data analysis and management, quantitative image quality assessment and modeling and optimization of data of any kind.

Email: akmishra@engmail.uwaterloo.ca

Supervisors

David Clausi, Paul Fieguth

Research topics

Biomedical Imaging
Computer Vision
Evolutionary Deep Intelligence
Image Segmentation/Classification
Remote Sensing
Video Analysis

Research demos

Decoupled Active Contours
Image Denoising

Publications

Journal Articles / Conference Papers

Journal Articles

Conference Papers

Top

Bi Hongbo

Bi Hongbo — Thu, 23 Mar 2023 17:17:57 +0000

I am a visiting researcher in the Vision and Image Processing Lab. I collaborate with professor David Clausi, professor Paul Fieguth, Post Doctoral Fellow (PDF) Christian Scharfenberger and Ph.D. Ahmed Gawish in the areas of image processing and digital watermarking.

Supervisors

David Clausi, Paul Fieguth

Research interests

My research interests are in the fields of digital watermarking, signal processing, pattern recognition and related fields. I am currently particularly interested in statistical modeling and feature extraction of natural images.

Research topics

Computer Vision
Multiresolution Techniques
Stochastic Models

Research demos

Action Recognition in Video
Compressed Sensing
Decoupled Active Contours
Statistical Textural Distinctiveness for Salient Region Detection in Natural Images

Publications

Journal Articles / Conference Papers

Journal Articles

Conference Papers

Top

Decoupled Active Contours

vipadmin — Mon, 20 Mar 2023 20:34:25 +0000

The accurate detection of object boundaries via active contours is an ongoing research topic in computer vision. Most active contours converge towards some desired contour by minimizing a sum of internal (prior) and external (image measurement) energy terms. Such an approach is elegant, but suffers from a slow convergence rate and frequently mis-converges in the presence of noise or complex contours. To address these limitations, a decoupled active contour (DAC) is developed which applies the two energy terms separately. Essentially, the DAC consists of a measurement update step, employing a Hidden Markov Model (HMM) and Viterbi search, and then a separate prior step, which modifies the updated curve based on the relative strengths of the measurement uncertainty and the non-stationary prior. By separating the measurement and prior steps, the algorithm is less likely to mis-converge; furthermore, the use of a Viterbi optimizer allows the method to converge far more rapidly than energy-based iterative solvers. The results clearly demonstrate that the proposed approach is robust to noise, can capture regions of very high curvature, and exhibits limited dependence on contour initialization or parameter settings. Compared to five other published methods and across many image sets, the DAC is found to be faster with better or comparable segmentation accuracy.

Finding the visual boundary using the Viterbi search

The “visual boundary estimation” problem is formulated as a shortest path finding puzzle on a discrete trellis. Viterbi search algorithm is employed to find the shortest path.

Circular trellis with an initial snake (blue circle), one possible, albeit unlikely snake (magenta), normals (green lines) and nodes (small circles along the normals)

Generation of non-stationary prior via importance sampling

To capture high curvature more samples are placed near high curvature to lower the penalty on curvature. The curvature guided importance sampling method is demonstrated below:

Statistical data fusion

The non-stationary prior is fused with measured boundary using a Linear Baysian estimation formulation. The effect measurement uncertainties is demonstrated below:

Downloads

Decouple Active Contour for Boundary Detection.pdf

Related people

Directors

Students

Alumni

Related research areas

Image Segmentation/Classification

Paul Fieguth

Paul Fieguth — Thu, 23 Feb 2023 20:12:37 +0000

Paul Fieguth received his B.A.Sc. degree from the University of Waterloo, Ontario, Canada, in 1991 and his Ph.D. degree from the Massachusetts Institute of Technology, Cambridge, in 1995, both degrees in electrical engineering. He joined the faculty at the University of Waterloo in 1996, where he is currently Professor and, since 2010, Department Chair in Systems Design Engineering. He has held visiting appointments at the University of Heidelberg in Germany, at INRIA/Sophia in France, at the Cambridge Research Laboratory in Boston, at Oxford University and the Rutherford Appleton Laboratory in England, and with postdoctoral positions in Computer Science at the University of Toronto and in Information and Decision Systems at MIT.

Students

Supervision – Current

PDF

ph.D.

M.A.Sc.

Supervision – Completed

PDF

ph.D.

M.A.Sc.

Research topics

Biomedical Imaging
BIOSCAN Insect Biodiversity Assessment
Computer Vision
Image Segmentation/Classification
Multiresolution Techniques
Remote Sensing
Scientific Imaging
Stochastic Models
Video Analysis

Research demos

3D Reconstruction of Underwater Scenes
Decoupled Active Contours
Image Denoising
Porous Media
SAR Sea Ice Image Synthesis
Skin Cancer Detection
Texture Classification

Publications

Journal Articles / Conference Papers

Journal Articles

Conference Papers

Top

David Clausi

David A Clausi — Mon, 13 Feb 2023 20:18:05 +0000

I completed my Ph.D. in Systems Design Engineering in 1996 and immediately afterwards, I worked in software medical imaging field at Agfa (nee Mitra) in Waterloo, Ontario. I started my academic career in 1997 as an Assistant Professor in Geomatics Engineering at the University of Calgary, Canada. In 1999, I returned to my alma mater and I am now a Professor specializing in the fields of Intelligent and Environmental Systems. In addition, I was the Associate Chair – Graduate Studies (2009-2012). I was the Co-chair of IAPR Technical Committee 7 – Remote Sensing during 2004-2006. I have numerous recognitions, including received five Outstanding Performance Awards, three Distinguished Performance Awards, the Sanford Fleming Teaching Excellence Award, and a University of Calgary Teaching Excellence Award. In 2010, I received the award for “Research Excellence and Service to the Research Community” by the Canadian Image Processing and Pattern Recognition Society (CIPPRS). In 2012, I received the Engineering Research Excellence Award.

Currently, I am the Associate Dean – Research & External Partnerships within the Faculty of Engineering, responsible for all research related activities in the largest Engineering school in Canada.

Research interests

My research interests are in the fields of computer vision, image processing, and pattern recognition with an emphasis on the automated interpretation of SAR (synthetic aperture radar) and sports analytics (ice hockey). I am an active interdisciplinary and multidisciplinary researcher, publishing refereed journal and conference papers in the diverse fields of remote sensing, sports analytics, computer vision, algorithm design, and biomechanics. My research efforts have led to successful commercial implementations including creating and selling a startup company (CREZ).