A colour histogram clustering technique for tissue analysis of healing skin wounds

A method for automatically quantifying the proportions of healthy granulation tissue in a healing skin wound using true colour image processing is described. A novel technique is proposed, involving the computation of the 3D RGB histogram of colour wound images and operations on it. Clustering of bins in the colour histogram has been used to achieve automatic segmentation of the different tissue types within the wound site: granulation tissue, slough and necrotic tissue. The technique makes use of the colour information in terms of RGB co-ordinates, rather than treating the R, G and B components separately.