Data Management for Live Plant Identification

Abstract

Plant identification by computers will surely find a wide range of applications including plant resource survey, plant data management, and education on plant taxonomy. Since computerized plant identification is a very challenging computer vision problem, the research and development in this field is still in its infancy. In this chapter, we first address the necessary and general issues regarding computer-aided plant identification and the management of a living plant database. We then describe a general approach that botanists adopt for plant identification, and typical systems for living plant identification and plant data management. This is followed by a discussion on a sophisticated approach for computer-aided plant identification supported by image processing, intelligent information processing, and plant data management systems. In this chapter, we also discuss various algorithms for processing leaf, flower and plant images, including image segmentation, leaf venation extraction, and flower region localization. In the discussion, we stress the unique features of plant images and special treatments needed for processing these images. Also discussed in this chapter are the feature extraction of leaf and flower images and the use of these features for leaf and flower image retrieval, a very important sub-task for computer-aided living plant identification. Finally, concluding remarks are made at the end of the chapter.