d="par0005">There is a potential for a completely unsupervised mapping of urban vegetation cover from an IKONOS-derived vegetation Index.
dd><dt class="label">•dt><dd>d="par0010">Two unsupervised techniques are investigated for extraction of urban trees/shrubs and ground vegetation cover types.
dd><dt class="label">•dt><dd>d="par0015">Techniques investigated include the Jenks natural breaks classification and an innovative two-step clustering technique.
dd><dt class="label">•dt><dd>d="par0020">The unsupervised techniques are simple, efficient and can be easily automated for use in low resource areas.
dd>dl>