Reef rubble representsthe broken components of the coraland reefstructure which could be in the form of dead,broken or other fragmented coral.The process to estimate the distribution of reef rubble is currently done manually and thus takesa long timeto completeand is laborious. This paper presentsan image-processing-basedmethod to estimate the distribution of reef rubbles in a coral reef environmentfrom a still image. The method is basically a series of image processing steps includingimage complement, image binarization, edgedetection, smoothing by Weiner filter and followed by erosion and dilation process.The experimentalresults showedthat the system wasable to roughly estimate the distribution of reef rubble.
At present, marine scientists employ manual method to estimate the components in coral reef environment,where Coral Point Count with Excel extensions (CPCe) software is used to determine the coral reef components and substrate coverage. This manual processis laboriousand time consuming,and needsexpertsto conduct the survey. In this paper, a prototype for estimating the distribution of sand cover in coral reef environment from still images by using colourextraction methods was introduced. The coloursegmentation called delta E was used to calculate the colourdifference between two coloursamples. Another method used wascolourthresholdby setting the range of sand colourpixels. Thesystem was developed by using a MATLAB software withimage processing toolbox. The developed system was semi-automatic computer-based system that can be used by researcherseven with little knowledge and experience to estimatethepercentage of sand coveragein coral reef still images.