Image processing for smart browsing of ocean colour data products and subsequent incorporation into a multi-modal sensing framework

摘要

Ocean colour is defined as the water hue due to the presence of tiny plants containing the pigment chlorophyll, sediments and coloured dissolved organic material and so water colour can provide valuable information on coastal ecosystems. The ‘Ocean Colour project’ collects data from various satellites (e.g. MERIS, MODIS) and makes this data available online. One method of searching the Ocean Colour project data is to visually browse level 1 and level 2 data. Users can search via location (regions), time and data type. They are presented with images which cover chlorophyll, quasi-true colour and sea surface temperature (11 μ) and links to the source data. However it is often preferable for users to search such a complex and large dataset by event and analyse the distribution of colour in an image before examination of the source data. This will allow users to browse and search ocean colour data more efficiently and to include this information more seamlessly into a framework that incorporates sensor information from a variety of modalities. This paper presents a system for more efficient management and analysis of ocean colour data and suggests how this information can be incorporated into a multi-modal sensing framework for a smarter, more adaptive environmental sensor network.