Developing an Automated Identification Engine using Cluster Computing Techniques to Identify Objects in Imagery

摘要

Machine vision algorithms have been applied in a wide range of scientific applications. We apply machine vision techniques to provide an analytical tool to automatically identify biological organisms from digital imagery. The decline of global biodiversity caused by the over-exploitation of natural resources, land use change, and climate change has been well established and the need for powerful analytical computational tools is urgent. Estimates suggest that described species represent less than 15 per cent of the potential number of organisms on the earth. We have developed a general purpose machine vision pattern classifier that can be applied to any form of digital imagery but which has been specifically used to identify organisms from imagery. DAISY (the Digital Automated Identification SYstem) is a modular system which was developed with the P3M cluster programming environment.