Midsummer Research Picks 2015

Cuckoo to you

How do you colour a map so that no two regions that share a border are the same colour, what is the minimum number of colours needed? Similarly, how do you colour, or tag a network so that no adjacent nodes have the same colour or tag? This is the classic “graph colouring problem” one of a group of difficult puzzles mathematicians call NP-hard problems. Countless approaches have attempted to solve the “GCP”. Now, a team in Algeria has turned to quantum theory and a statistical approach known as a cuckoo search algorithm to make their attempt to surmount this mountainous task. The approach can with relative speed search for solutions for a given graph and the final results, the team says are “very encouraging”, offering not only a possible answer to an old problem but demonstrating how a quantum computer might carry out such difficult tasks.

Djelloul, H., Layeb, A. and Chikhi, S. (2015) ‘Quantum inspired cuckoo search algorithm for graph colouring problem’, Int. J. Bio-Inspired Computation, Vol. 7, No. 3, pp.183–194.

Can you see clearly now?

A new approach to testing whether or not a “de-fogging” algorithm can truly enhance a photo or other images with quality compromised by bad weather or degradation, without loss of too much information, has been developed by a team in China. The team has first devised a method for generating a synthetic fog so that the defogging algorithms can be evaluated and the secondly they have developed a way to assess the results of defogging software that corresponds closely to visual assessment without relying on a reference. Automated image manipulation and enhancement is becoming increasingly important in science, security and other areas and so being able to judge objectively whether or not a given defogging algorithm will work with a given class of image is also increasingly important so that automated enhancement might be implemented.

Guo, F., Tang, J. and Cai, Z. (2015) ‘An objective assessment method for image defogging effects’, Int. J. Autonomous and Adaptive Communications Systems, Vol. 8, Nos. 2/3, pp.180–199.

Alluring and luring women entrepreneurs

Researchers in Mauritius suggest that female entrepreneurship is a significant component of economic growth and business innovation in the modern world. However, they suspect that women still face many challenges that their male counterparts do not. Moreover, they also suspect that there remains a gap between the successes men and women have in different areas of entrepreneurship. They have investigated the state of play in Mauritius and found that the gap between male and female entrepreneurs remains a fundamental problem. They add that despite the equity in what the authorities and private institutions offer, there are, in reality, subtle and disarming differences between how male and female entrepreneurs are treated.

Mahadeo, J.D., Dusoye, I.C. and Aujayeb-Rogbeer, A. (2015) ‘Women and entrepreneurship: an alluring or luring option’, Int. J. Entrepreneurship and Small Business, Vol. 25, No. 3, pp.351–374.

A fruity spot the difference

As humans, we can distinguish very easily between a potato and a mango just from a quick glance, but in automated logistics for food selection, transport and sales, there is a need to develop computer vision systems that can distinguish between different fruit and vegetables. Researchers in India have introduced a framework for analysing images that involves background subtraction, feature extraction, training and classification so that colour and texture can be distinguished and similarly shaped fruit and vegetables identified. The team says that their experimental results show that the proposed approach supports accurate fruit and vegetable recognition and performs better than standalone colour and texture features.

Dubey, S.R. and Jalal, A.S. (2015) ‘Fruit and vegetable recognition by fusing colour and texture features of the image using machine learning’, Int. J. Applied Pattern Recognition, Vol. 2, No. 2, pp.160–181.

Author: David Bradley

Award-winning, freelance science writer based in Cambridge, England.