The fungus Mycosphaerella (or mycos) causes internal fruit rot to cucumbers which is not visible on the outside during the first stage. During sorting the contaminated cucumbers are difficult to recognize, even by experienced personnel who separates these as good as possible. Due to sorting on mycosphaerella the processing speed must be reduced often. Cucumbers which are taken out have no monetary value and are disposed as waste . Contaminated cucumbers that are still sold to retail and reach consumers in the end. This "rotten" cucumbers are highly undesirable because they are not saleable and the "healthy" cucumbers may be infected. Difficult to quantify and therefore perhaps even more serious are the cost of declining sales due to affected cucumbers on the shelves and the chance of declining sales of cucumber in general.
People who are involved estimate the damage at € 1.10 per meter a year. Regarding an area of 620 hectares in the Netherlands, this represents a loss of more than 6 million euros for the production chain. For this reason, grower associations and retail would like to see infected cucumbers removed prematurely from the chain.
Despite several studies on the causes of mycosphaerella it remains an important recurring problem. This year, Wageningen UR and Kompany started a project with the aim to develop a method for automatically detecting and removing mycosphaerella infected cucumbers during sorting. This should lead to a technical solution for the cucumber sector which must be financially viable in terms of investments. The central question is whether it is possible to detect mycos on the basis of ( hyper) spectral image processing techniques (transmission or reflection) combined with geometric ( colour, texture and shape ) image analysis techniques. The study builds on the results achieved in a previous project where infrared transmission is used as detection method.