Computational Challenges in Image Processing

Background

Image processing is a dynamic and fast moving field of research. Recent advances in the area have led to an explosion in the use of images in a variety of scientific and engineering applications. New approaches are constantly being developed by mathematicians, engineers and computer scientists to be applied to image processing problems. Image processing, along with mathematical imaging and computer vision have become fundamental for gaining information on various aspects in medicine, the sciences, and technology, in the public and private sector equally. The rapid development of new imaging hardware, the advance in medical imaging, the advent of multi-sensor data fusion and multimodal imaging, as well as the advances in computer vision have sparked numerous research endeavours leading to highly sophisticated and rigorous mathematical models and theories.

There are many computational challenges in image processing. These include issues such as the handling of image uncertainties that cannot be otherwise eliminated, including various sorts of information that is incomplete, noisy, imprecise, fragmentary, not fully reliable, vague, contradictory, deficient, and overloading. However, some computational techniques such as fuzzy logic, neural networks, and evolutionary methods have shown great potential to solve such image processing problems.

This afternoon workshop, part of the Isaac Newton Institute Research Programme Variational Methods and Effective Algorithms for Imaging and Vision, brought together mathematicians, computer scientists and engineers from both the research and industry communities. Talks from academics and end-users explored various computational challenges around areas of image processing.

Aims and Objectives

This Open for Business workshop aimed to extend the reach of the Isaac Newton Institute research programme, by fostering exchange between different groups of researchers and practitioners who are involved in imaging science. The event highlighted both some of the challenges and potential novel solutions for computational image processing. Talks and discussion highlighted possible new mathematical models which are needed to address the ever growing challenges in applications and technology, generating new demands that cannot be met by existing mathematical concepts and algorithms.

The Programme of talks featured academic state-of-the-art talks, as well as end-user challenge type presentations and included areas such as:

Variational image processing

Optimisation and machine learning approaches

Development of computational methods that enable semiautomatic analysis

Algorithmic challenges of global optimisation and convex relaxation methods as well as stochastic optimisation for large- and high-dimensional imaging problems

Dynamic image processing challenges

Medical image processing challenges

Remote sensing: aerial and satellite image interpretations and Image processing challenges

The workshop included a poster exhibition, which ran during the lunch and the drinks/networking session. It brought together industrial and academic experts from a diverse set of backgrounds in mathematics, computer science and information engineering. Many relevant sectors included computer and software engineering, medical and biomedical, security/biometrics, environmental monitoring, industrial automation/inspection, traffic management, media and creative industries.